Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Empiric risk to first degree relatives of individuals with non-autosomal dominant alzheimer’s disease Bourque, Sylvie Alice 1997

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_1997-0381.pdf [ 5.3MB ]
Metadata
JSON: 831-1.0087775.json
JSON-LD: 831-1.0087775-ld.json
RDF/XML (Pretty): 831-1.0087775-rdf.xml
RDF/JSON: 831-1.0087775-rdf.json
Turtle: 831-1.0087775-turtle.txt
N-Triples: 831-1.0087775-rdf-ntriples.txt
Original Record: 831-1.0087775-source.json
Full Text
831-1.0087775-fulltext.txt
Citation
831-1.0087775.ris

Full Text

EMPIRIC RISK TO FIRST DEGREE RELATIVES OF INDIVIDUALS WITH NON-AUTOSOMAL DOMINANT ALZHEIMER'S DISEASE BY SYLVIE ALICE BOURQUE B.Sc. L'Universite de Moncton, 1994 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDIES (Department of Medical Genetics) We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA July 1997 © Sylvie Alice Bourque, 1997 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department The University of British Columbia Vancouver, Canada DE-6 (2/88) Abstract The primary objective of this thesis was to determine if patients with early onset (<65 years) non-autosomal dominant AD are more genetically loaded i.e., have more AD liability genes, than patients with late onset (> 65 years) non-autosomal dominant AD. Secondary aims were: (i) to examine the effect(s) of the gender of the index case on the risk of AD to first degree relatives (ii) to examine the effect(s) of the gender of the first degree relative on their AD risk (iii) to examine the effect(s) of the relationship between the index case and the first degree relative (parent or sib) on the relative's risk of AD (iv) to compare the risk estimates for first degree relatives of non-autosomal dominant AD cases with those for first degree relatives of controls to determine genetic loading, and (v) if certain subgroups of first degree relatives of AD cases were at an increased risk, to determine whether this was reflected in the general population. Kaplan-Meier age-specific risk estimates were calculated for the 2519 first degree relatives of the 453 early and late onset non-autosomal dominant AD cases seen at the Clinic for Alzheimer Disease and Related Disorders -UBC Site from 1985 to 1995. Kaplan-Meier risk estimates were also calculated for 4324 first degree relatives of 796 controls taken from the Canadian Study of Health and Aging (CSHA). The cumulative lifetime risk estimate to age 88 for first degree relatives of early onset cases was 7.9± 2.1% compared to 8.0± 1.5% for first degree relatives of late onset cases, and 4.1 ± 0.6% for first degree relatives of controls. The difference between the cumulative lifetime risk estimates for first degree relatives of early and late onset cases was not significant (Z=0.04, p=0.48). The cumulative lifetime risk estimate for first degree relatives of late onset cases was significantly different from the cumulative lifetime risk estimate for first degree relatives of controls (Z=2.41, p<.01). These results argue against increased genetic loading in early onset non-autosomal dominant AD ii cases in relation to late onset non-autosomal AD cases. Increased familial aggregation in first degree relatives of late onset non-autosomal dominant cases in relation to controls suggests that genetic loading in non-autosomal dominant AD cases may be greater than in controls, and therefore provides data that there might be genetic factors contributing to liability for non-autosomal dominant AD. Risk estimates for parents and sibs and for first degree relatives of female and male cases and controls were not significantly different. The risk to female first degree relatives of cases was significantly higher than the risk to male first degree relatives; however this was shown to be reflect the gender ratio found in the general Canadian population. iii Table of Contents Abstract ii Table of Contents iv List of Tables viii List of Figures x Acknowledgments xii 1. Introduction 01 1.1 Background 01 1.2 Diagnosis 03 1.3 Etiology of Alzheimer's Disease 06 1.3.1 Non-Genetic Risk Factors 06 1.3.1.1 Gender 06 1.3.1.2 Exposure to Aluminum 07 1.3.1.3 History of Head Injury 08 1.3.2 Non-Genetic Protective Factors 09 1.3.2.1 Estrogen Supplements 09 1.3.2.2 Non-Steroidal Anti-Inflammatory Drugs 10 1.3.2.3 Education 11 1.3.2.4 Smoking 11 1.4 Genetics of Alzheimer's Disease 12 1.4.1 B-Amyloid 12 1.4.1.1 B-Amyloid Protein Proteolysis Pathways 13 1.4.1.2 BAPP and Early Onset Familial Autosomal Dominant AD 15 1.4.1.3 The APP717 Mutations 15 1.4.1.4 The Swedish Double Mutation- APP670/671 17 1.4.1.5 The Flemish and Dutch Mutations 18 iv 1.4.1.6 Possible Mechanisms of Pathogenesis 18 1.4.2 The Presinilin Genes: PS-1 and PS-2 19 1.4.2.1 Expression Patterns 22 1.4.2.2 Age of Onset 23 1.4.2.3 Possible Functions of the Presinilins 23 1.4.2.4 PS-1 and Late Onset Alzheimer's Disease 24 1.4.3 Apolipoprotein E 25 1.4.3.1 ApoE and Alzheimer's Disease 26 1.5 Transgenic Models 30 1.6 Empiric Risk Estimate Studies in Alzheimer's Disease 31 1.7 Non-Autosomal Dominant Alzheimer's Disease 32 1.8 Clinic for Alzheimer Disease and Related Disorders, University Hospital- UBC Site 34 1.9 The Canadian Study on Health and Aging 34 2. Objectives of the Study 37 3. Relevance of the Study 38 3.1 The Significance of Alzheimer's Disease in the Canadian Population 38 3.2 Risk Assessment for Relatives of Alzheimer's Disease Cases 39 4. Material and Methods 41 4.1 Cases 41 4.1.1 Relatives of Cases 44 4.2 Controls 45 4.2.1 Relatives of Controls 45 4.3 Kaplan-Meier Age-Specific Risk Estimates 46 4.4 Cox Proportional Hazard Ratios 50 4.5 Correction for the Gender Ratio in the General Elderly Population 52 v 5. Results 53 5.1 Samples Used for Analysis 53 5.1.1 Cases 53 5.1.2 Controls 53 5.2 Risk Estimates for First Degree Relatives of Early Onset Cases, Late Onset Cases and Controls 56 5.3 Risk Estimates for First Degree Relatives by Gender of the Relatives 59 5.3.1 Risk Estimates for Female and Male First Degree Relatives of Early Onset Cases 59 5.3.2 Risk Estimates for Female and Male First Degree Relatives of Late Onset Cases 59 5.3.3 Risk Estimates for Female and Male First Degree Relatives of Controls 64 5.4 Risk Estimates for First Degree Relatives by Relationship to the Index Case 64 5.4.1 Risk Estimates for Parents and Sibs of Early Onset Cases 64 5.4.2 Risk Estimates for Parents and Sibs of Late Onset Cases 64 5.4.3 Risk Estimates for Parents and Sibs of Controls 71 5.5 Risk Estimates for First Degree Relatives by Gender of the Proband 71 5.5.1 Risk Estimates for First Degree Relatives of Female and Male Early Onset Cases 71 5.5.2 Risk Estimates for First Degree Relatives of Female and Male Late Onset Cases 76 5.5.3 Risk Estimates for First Degree Relatives of Female and Male Controls 76 5.6 Summary of Risk Estimates for Various Subgroups of First Degree Relatives 81 5.7 Relative Risk Odds Ratios for First Degree Relatives of Early Onset Cases and Late Onset Cases by Gender, and Relationship to Index Case 82 5.8 Relative Risk Odds Ratios for First Degree Relatives of Controls by Gender, and Relationship to Index Case 83 vi 5.9 Relative Risk Odds Ratios for First Degree Relatives of Early Onset Cases, Late Onset Cases and Controls by Gender, and Relationship to the Index Case 85 5.10 Risk Estimates for Female and Male First Degree Relatives of All Cases with and without Corrections for the Gender Ratio in the General Population 87 6. Discussion 94 6.1 The Difference in Genetic Loading in Early Onset Non-Autosomal Dominant Cases and Late Onset Non-Autosomal Dominant Cases of Alzheimer's Disease 94 6.2 Genetic Loading in Early Onset and Late Onset Cases Compared to Controls 96 6.3 Gender of the First Degree Relative 96 6.4 Relationship 99 6.5 Gender of the Index Case 100 6.6 The Influence of Genetic Factors in Non-Autosomal Dominant Cases of Alzheimer's Disease 100 6.7 Strengths and Weaknesses of our Study 104 7. Summary 106 Bibliography 108 Appendix A 122 Appendix B 126 vii List of Tables Table 1.1 Apolipoprotein E Polymorphisms 25 Table 1.2 The Effects of ApoE Genotypes on Risk and Age of AD Onset 28 Table 5.1 Diagnosis for 1311 Alzheimer Clinic Patients Seen at the Alzheimer Clinic from 1985 to 1995, After Evaluation 54 Table 5.2 Number and Percent of Affected First Degree Relatives of Early and Late Onset Cases by Gender and Relation to Index Cases 54 Table 5.3 Number and Percent of Affected First Degree Relatives of Controls, by Gender and Relationship to Controls 55 Table 5.4 Early Onset Cases, Late Onset Cases, and Controls, by Gender, Mean Age, and Mean Age of Onset of Dementia 55 Table 5.5 Age-Specific Risks for First Degree Relatives of Early Onset Cases, Late Onset Cases, and Controls 57 Table 5.6 Age-Specific Risks for Female and Male First Degree Relatives of Early Onset Cases 60 Table 5.7 Age-Specific Risks for Female and Male First Degree Relatives of Late Onset Cases 62 Table 5.8 Age-Specific Risks for Female and Male First Degree Relatives of Controls 65 Table 5.9 Age-Specific Risks for Sibs and Parents of Early Onset Cases 67 Table 5.10 Age-Specific Risks for Sibs and Parents of Late Onset Cases 69 Table 5.11 Age-Specific Risks for Sibs and Parents of Controls 72 Table 5.12 Age-Specific Risks for First Degree Relatives of Female and Male Early Onset Cases 74 Table 5.13 Age-Specific Risks for First Degree Relatives of Female and Male Late Onset Cases 77 Table 5.14 Age-Specific Risks for First Degree Relatives of Female and Male Controls 79 Table 5.15 Odds Ratios for Various Subgroups of First Degree Relatives of AD Cases 82 Table 5.16 Odds Ratio for Female First Degree Relatives of AD Cases (Reduced Model)83 viii Table 5.17 Odds Ratios for Various Subgroups of First Degree Relatives of Controls 84 Table 5.18 Odds Ratio for Female First Degree Relatives of Controls (Reduced Model) 84 Table 5.19 Odds Ratios for Various Subgroups of First Degree Relatives of AD Cases and Controls 86 Table 5.20 Odds Ratios for Various Subgroups of First Degree Relatives of AD Cases and Controls (Reduced Model) 86 Table 5.21 Age-Specific Risks for Female and Male First Degree Relatives of Index Cases 88 ix List of Figures Figure 1.1 Most Common Subtypes of Dementia in the Canadian Population 2 Figure 1.2 B Amyloid Precursor Protein 14 Figure 1.3 Structure of the Presinilin Proteins 20 Figure 3.1 Proportion of AD Cases Caused by Mutations: Sporadic and Famililal Cases Combined 39 Figure 4.1 Autosomal Dominant Pedigrees 43 Figure 5.1 Age-Specific Risk (%) to First Degree Relatives of Early Onset Cases, Late Onset Cases, and Controls 58 Figure 5.2 Age-Specific Risk (%) to Female and Male First Degree Relatives of Early Onset Cases 61 Figure 5.3 Age-Specific Risk (%) to Female and Male First Degree Relatives of Late Onset Cases 63 Figure 5.4 Age-Specific Risk (%) to Female and Male First Degree Relatives of Controls 66 Figure 5.5 Age-Specific Risk (%) to Sibs and Parents of Early Onset Cases 68 Figure 5.6 Age-Specific Risk (%) to Sibs and Parents of Late Onset Cases 70 Figure 5.7 Age-Specific Risk (%) to Sibs and Parents of Controls 73 Figure 5.8 Age-Specific Risk (%) to First Degree Relatives of Female and Male Early Onset Cases 75 Figure 5.9 Age-Specific Risk (%) to First Degree Relatives of Female and Male Late Onset Cases 78 Figure 5.10 Age-Specific Risk (%) to First Degree Relatives of Female and Male Controls 80 Figure 5.11 Age-Specific Risk (%) to Female and Male First Degree Relatives of Cases 89 Figure 5.12 Age-Specific Risk (%) to Female and Male First Degree Relatives of Cases Adjusted for a Gender Ratio (F:M) of 1.2:1 90 Figure 5.13 Age-Specific Risk (%) to Female and Male First Degree Relatives of Cases Adjusted for a Gender Ratio (F:M) of 1.5:1 91 x Figure 6.1 Age-Specific Risk (%) to First Degree Relatives of All Cases, and Non-Autosomal Dominant Cases xi Acknowledgments I would like to thank my thesis committee for their support and comments: Dr. B.L. Beattie, Dr. M. Harris, Dr. S. Langlois, and Dr. W. Robinson. I would like to thank my defense committee: Dr. J. Friedman, Dr. B. McGillivray, and Dr. W. Robinson. I would like to thank Irene Yee, without whose help and guidance this thesis would not have been complete. I would like to thank Dr. A. D. Sadovnick for her guidance and support throughout the last three years, and acknowledge that without her assistance this work would not have been possible. xii 1. Introduction 1.1 Background The term dementia refers to a group of diseases characterized by a progressive and usually irreversible decline of mental functions. Symptoms of dementia include memory loss, disorientation, cognitive decline and inappropriate social behavior. A diagnosis of dementia requires an unmistakable deterioration in two cognitive domains relative to the patient's previous level of function (American Psychiatric Association 1994). This is determined by a history of intellectual decline and must be documented by formal mental status testing with neuropsychological tests such as the Mini Mental State (MMS) (Folstein 1975) and the Modified Mini Mental State (3MS) exams (Teng & Chui, 1987). Several types of dementia exist. Therefore, once a diagnosis of dementia is made, it is possible to differentiate the type of dementia by its etiology, e.g. dietary deficiencies (vitamin B12, niacin, folic acid), occurrence of central nervous system conditions (Alzheimer's dementia, Pick's dementia, and Huntington's disease), and substance abuse (alcohol). Some dementias are treatable and reversible, at least to some degree. These include dementia related to dietary deficiencies, substance abuse or depression. Other dementias are irreversible (e.g. Alzheimer's dementia, Huntington's disease, Pick's dementia). Critical to the differential diagnosis of dementia is the recognition that cognitive functions are not lost at the same time or at the same rate in these various sub-types of dementia. Therefore, it is the relative conservation and loss of different cognitive functions that are used to distinguish subtypes of dementia. Alzheimer's disease (AD) is a neurodegenerative disorder of the central nervous system (brain) which causes progressive memory and cognitive decline during mid- to late-adult life. AD is the most common type of dementia in developed nations where it accounts for approximately two thirds of all dementia cases (CSHA(a) 1994)(see figure 1.1). There is 1 a preponderance of female cases of AD, with a female:male prevalence ratio of 1.5:1 (CSHA 1994 (a)). Figure 1.1: Most Common Subtypes of Dementia in the Canadian Population3 Parkinson's Disease oo/ Other Causes The onset of dementia of the Alzheimer type is usually gradual (i.e. not sudden) and involves a progressive cognitive decline rather than a step-wise progression. Nevertheless, a pattern of sudden decline followed by stabilization or a step-wise progression of cognitive decline does not rule out a diagnosis of AD. Various patterns of cognitive deficits can be seen, but the most common pattern is one of an insidious onset, with early deficits in recent memory followed by the development over several years of aphasia (language impairment), apraxia (impaired ability to carry out motor activities despite intact motor function), and agnosia (failure to recognize or identify objects despite intact sensory function) (American Psychiatric Association 1994). a Data taken from CHS A 1994(a) 2 The initial signs and symptoms of AD are often difficulties in completing instrumental activities of daily living. Initially, an individual with AD may become repetitive and/or may forget daily events such as phone messages and appointments. Basic financial management may become problematic. Bills are often left unpaid or, alternatively, may get paid twice. Persons may experience difficulty with orientation in unfamiliar surroundings and may, of necessity, restrict driving to familiar areas. Gradually, the individual may develop exaggerated memory aids such as written reminders. Persons in the early stages of AD may function reasonably well in familiar surroundings and/or using memory aids. It is not unusual for children to only become aware of a parent's difficulty (or severity of the problem) after the death of a spouse. Over time, as AD progresses, the individual loses the ability to complete most, if not all, activities of daily living. The average life expectancy for AD patients is 9 years from the time of symptom onset, however some patients have survived for up to 23 years (Bird et al., 1989). 1.2 Diagnosis In the Diagnostic and Statistical Manual of Mental Disorders, Revised third edition (DSM-III-R) (American Psychiatric Association 1987), dementia is described as impairment in short term (difficulty in learning new information) and long-term memory (difficulty remembering past personal information or facts of common knowledge) associated with at least one of the following: i) impairment in abstract thinking (trouble coping with novel tasks), ii) impaired judgment (disregard for the conventional rules of social conduct), iii) other disturbances of higher cortical function (aphasia (language impairment), constructional ability (ability to reconstruct three dimensional figures), agnosia (failure to recognize or identify objects despite intact sensory function), and apraxia (inability to carry out motor activities despite intact comprehension and motor function)), or iv) personality 3 changes (an alteration or accentuation of premorbid traits) that significantly interferes with work or usual social activities or relationships with others. In DSM-IV (American Psychiatric Association 1994), the definition of dementia was modified to include the following criterion: the cognitive deficits must be severe enough to cause significant impairment in social or occupational functioning and must represent a decline from a previous level of functioning. Once a diagnosis of dementia is made, the definitive diagnosis of dementia of the Alzheimer's type can be made only after neuropathological evidence is obtained from a brain autopsy (or brain biopsy). Neuropathologically, AD is characterized by an abundant presence of extracellular amyloid plaques, composed of an accumulation of a 39-42 amino acid B-amyloid peptide, and intracellular neuronal neurofibrillary tangles. Neurofibrillary tangles result when tau protein, associated with microtubule assembly and stabilization in neurons, becomes hyperphosphorylated, and loses its ability to bind to microtubule segments. As a result, microtubules, which are responsible for cell structure (neuronal cytoskeleton) become unstable, and bundles of paired helical filaments (PHF'S) (tau protein carrying more phosphate groups than normally and at sites not ordinarily phosphorylated) aggregate, thus giving the appearance of tangles (Lee et al., 1991). Other common histopathological changes that can be found are granulovascular degeneration, neuronal loss, astrocytic gliosis, and cerebral amyloid angiopathy. Although a definite diagnosis is based on pathological findings, diagnostic criteria exist for the clinical diagnosis of dementia and AD. It is possible to diagnose Dementia of the Alzheimer's type in living patients by a process of exclusion of all other causes of 4 dementia and subsequently confirming this diagnosis after death. Among the other causes of cognitive deficits to be ruled out are1: 1) Other central nervous system conditions that cause progressive deficits in memory and cognition (e.g. cerebrovascular disease, Parkinson's disease, Huntington's disease, subdural hematoma, normal pressure hydrocephalus, brain tumor). 2) Systemic conditions known to cause dementia (e.g. hypothyroidism, vitamin B12 or folic acid deficiency, niacin deficiency, hypercalcemia, neurosyphilis, HIV infection) 3) Substance-induced conditions In 1984, a Work Group on the diagnosis of Alzheimer's disease established by the National Institute of Neurological and Communicative Disorders and Stroke (NINCDS) and the Alzheimer's and Related Disorders Association (ADRDA) developed criteria for the clinical diagnosis of AD in response to the need for uniform diagnostic criteria for AD, especially for research purposes (McKhann et al., 1984). These criteria, widely used by researchers around the world, can have an inter-observer agreement of up to 95% (Forette et al., 1989; Baldereschi et al., 1994) and a clinicopathologic agreement of up to 96% (Tiemey et al., 1988; Kosusnen et al., 1996) in specialized centres. These criteria will be discussed in greater detail in section 4.1. 1 taken from DSM-IV (American Psychiatric Association 1994; pp142-143) 5 1.3 E t io logy of A l z h e i m e r ' s D i s e a s e The etiology of Alzheimer's disease is heterogeneous. A wide range of risk factors have been reported for AD, including genetic and non-genetic risk factors, as well as interactions between these two. The three major accepted risk factors are a positive family history of dementia (Li et al., 1992; Van Duijn et al 1994 (a); Fratiglioni et al., 1993), specific genetic mutations (see section 1.4), and advancing age. Other possible non-genetic risk factors for AD include: gender, head trauma, exposure to aluminum, and a low level of education. Protective factors have also been suggested and include: a history of estrogen supplement use, a history of non steroidal anti-inflammatory drug use, and smoking. 1.3.1 N o n - G e n e t i c R i s k Fac to rs 1.3.1.1 Gender The prevalence of AD is higher for women than for men (CSHA 1994(a)). In late onset families, families in which the average age of onset of affected members is above the age of 60 years, women have an average three-fold greater risk than men to develop AD (Payami et al., 1996). It is still unclear whether the observed female preponderance of AD is due to greater longevity in women (Corder et al., 1995), a sudden decrease in hormone levels at the onset of menopause, or some yet unknown factors. Recent studies have looked at the possibility that the e4 allele of the apolipoprotein E gene (Apo E) might exert a greater risk to women carriers than to men. Results show that although there are no differences in the mean age of onset between men and women e4 carriers, a large difference exists in the proportion of unaffected men carrying an e4 allele compared with the proportion of unaffected women carrying an e4 allele. It was shown that while 56% of women carrying one e4 allele developed AD at some point during their lifetime, only 35% 6 of men with one e4 allele were similarly affected (Payami et al., 1994; Yasuda et al., 1995). (The role of Apo E in AD will be discussed in more detail in section 1.4.3.) 1.3.1.2 Exposure to aluminum Aluminum has been detected in both senile plaques and neurofibrillary tangle bearing neurons in brains of AD patients. These findings have been interpreted as evidence that exposure to aluminum may be important in the causation of the disease. In some areas, aluminum sulfate is used as a flocculant in the treatment of water, as it helps in the removal of suspended matter and highly colored humic substances (Martyn et al., 1989). Results of studies looking for an association between aluminum concentrations in drinking water and the progression of Alzheimer's disease are very controversial. While some studies have shown a very clear positive association (Martyn et al 1989), others have failed to find any association (Forster et al., 1995). Recent reports (Forbes 1996) tried to explain these discrepancies by suggesting that future studies concentrate more on (i) late onset populations in which there would be a greater chance of aluminum accumulation due to a longer period of consumption, (ii) populations in areas of high water aluminum concentrations, and (iii) the concentration of other water constituents which might affect the nature of the association between soluble aluminum concentrations and the development of Alzheimer's disease. In a preliminary report, Taylor (1995) looked at the concentration of soluble silicon, which determines the bio-availability of all dietary exposures to soluble aluminum. He found a possible "protective" effect, i.e. when soluble silicone concentrations are low, soluble aluminum concentrations are high. Another possible source of exposure to aluminum is an occupation where aluminum dust can be transported to the brain via the nasal olfactory pathway (Salib & Hillier 1996). However, in a recent case-control study looking specifically for an association between 7 occupational exposure and development of Alzheimer's disease, no association was found (Salib & Hillier 1996). Other possible sources of aluminum are: aluminum containing antacids, antiperspirants containing aluminum, and tea; however all of these have now been discredited (CSHA 1994(b); Forster et al., 1995; Li et al., 1992) 1.3.1.3 History of Head Injury It has been hypothesized that head trauma might accelerate neuronal loss by increasing B-amyloid protein production in the brain, thereby accelerating the arrival to the "threshold" of neuronal loss that separates the cognitively normal individual from the cognitively impaired individual. Several studies have shown that patients with a history of head injury do indeed have an earlier onset of AD than do individuals without a history of head injury (Gedye et al., 1989; Sullivan et al., 1987). Nevertheless, head injury is neither necessary nor sufficient to cause AD . The risk of AD seems to be determined by the severity of head injury, as several studies have now confirmed that the risk of AD is increased in individuals with a history of severe head injury which includes loss of consciousness (Mayeux et al., 1993 & 1995; Sullivan et al., 1987; Mortimer et al., 1985), while a minor head injury, without loss of consciousness, seems insufficient to increase the risk (Jordan et al., 1990; Gedye et al., 1989). The time of occurrence of the head injury is also an important factor. Head injuries occurring later in life, after the age of 70 years seem to carry a greater risk (Mayeux et al., 1995). History of head trauma has recently proven to be a perfect example of non-genetic factors working in synergy with genetic factors to cause a disease. In keeping with the theory that increased B-amyloid production after head injury leads to an increase in neuronal death, Mayeux et al.(1995) showed that head injury increases the risk of developing AD only in individuals carrying the e4 allele of the apolipoprotein E (Apo E). 8 This is accentuated by the fact that Strittmatter and co-workers (1993) showed that the Apo E e4 allele aggregates B-amyloid at a quicker rate than do the other apoE alleles and that the Apo E E4 allele is associated with greater density of senile plaques in the brain parenchyma. Thus, individuals with one or more copies of e4 might be more likely to deposit p-amyloid in cortex following head injury than individuals with no e4 allele. B-amyloid deposition may result in cell death by binding to specific molecules on the cell surface or by rendering neurons more susceptible to excitotoxic, ischemic or other metabolic constituents. 1.3.2 N o n - G e n e t i c Protect ive F a c t o r s 1.3.2.1 Estrogen Supplements Estrogen replacement in post-menopausal women has been implicated as a possible protective factor for AD by: (i) increasing cerebral blood flow, (ii) reversing glucocorticoid damage, (iii) stimulating the secretase metabolism of the amyloid precursor protein, (iv) stimulating cholinergic markers, (v) preventing neuronal atrophy, or (vi) acting as an important co-factor in the actions of nerve growth factors (Burns & Murphy 1996). Recent studies (Tang et al., 1996) have shown that a history of estrogen use during the post menopausal period significantly delays the onset of AD and lowers the risk of AD. Additionally, it has been shown that duration of estrogen use and dosage may be important factors in risk reduction. It has been observed that women with a history of long term use and higher dosages have the lowest risk (Tang et al., 1996; Paganni-Hill & Henderson 1996). It has also been shown that women with AD who use estrogen supplements have better cognitive skills than demented women not receiving estrogen supplements (Henderson et al., 1994; Henderson et al., 1996; Schneider et al., 1996). 9 1.3.2.2 Non-Steroidai Anti-Inflammatory Drugs There is now a body of evidence that inflammation is an important contributor to both Alzheimer's disease pathology and the neurodegeneration that underlies Alzheimer dementia (reviewed by Rogers et al., 1996). However, it is thought that inflammation is unlikely to have an etiologic role in AD but is rather a reaction to a trigger. The hypothesis that non-steroidal anti-inflammatory drugs (NSAIDS) may have a protective effect(s) in AD is based on the observation that rheumatoid arthritis patients are at a reduced risk of developing AD (Myllykangas-luosujarvi & Isomaki 1994). This observation led to the hypothesis that AD is a chronic inflammatory condition, and NSAIDS, which are used to treat rheumatoid arthritis, are the risk altering factor. Evidence supporting the contribution of inflammation to the neurodegeneration underlying AD dementia is the significant elevation of inflammatory proteins such as interleukin 1, interleukin-6, and tumor necrosis factor in AD brains compared to non-demented elderly control brains (McGeer & Rogers, 1992). Acute phase reactants such as cc1-antichymotrypsin and cc2-macroglobulin, which are components of senile plaques, are also potent protease inhibitors, and thus their induction may influence the degradation of the amyloid precursor protein into B-amyloid (Reviewed by Rogers et al., 1996 and Aiesen & Davis 1994). Many epidemiologic studies have now shown that NSAIDS are protective against AD (CSHA 1994(b); Breitner et al., 1994; Andersen et al., 1995; Li etal., 1992). It has been shown that NSAIDS not only delay the onset of disease but also slow down the cognitive deterioration in patients already affected (Rich et al., 1995; Breitner et al., 1994). A clinical trial with the NSAID indomethacin in which 28 patients were randomly assigned to drug or placebo for six months showed that patients taking indomethacin improved by an average of 1.3% on cognitive status tests, while placebo treated patients deteriorated by an average of 8.4% (Rogers et al., 1993). The beneficial effects of indomethacin use were statistically 10 significant.which indicates that mild to moderate AD patients treated with indomethacin were in fact protected from the degree of cognitive decline exhibited by untreated, or placebo treated patients. Other clinical trials of NSAIDS are now underway. 1.3.2.3 Education It has been hypothesized that higher education contributes to an individual's "reserve" against dementia by supplying a set of skills or repertoires that allow an individual to cope longer before the clinical manifestations of AD emerge. Several studies have now shown that an increased level of education is associated with a lower risk of dementia (Stern et al., 1994; Bonaiuto et al., 1990; CSHA 1994(b); Plassman et al., 1995). 1.3.2.4 Smoking An inverse relationship between a history of smoking and AD has been suggested (Graves et al., 1991). However, this relationship might only be present in apoE e4 carrying individuals with a family history of dementia (VanDujin et al., 1995). The nicotine binding site density in the cerebral cortex of AD patients is reduced compared with controls, and this is more pronounced in individuals carrying at least one e4 allele. Studies have now shown that introduction of nicotine produces an increase in the number of nicotinic cholinergic recognition sites in the cerebral cortex (Benwelle et al., 1988). Therefore, it is possible that these two factors, having opposing effects on the regulation of nicotinic cholinergic recognition sites, might interact to produce an apparent protective effect against the onset of AD. However, it is important to note that the negative effects of smoking far outweigh the protective effect against AD, and therefore smoking should not be advocated (Shafer & Nett, 1995). 11 1.4 G e n e t i c s of A l z h e i m e r ' s d i s e a s e Three genes have now been identified in the cause of early onset AD, BAPP, PS-1 and PS-2. Six point mutations have been described in a total of 20 unrelated families in the B-amyloid precursor protein gene (BAPP). Mutations at position 717 of the gene account for 5% of all early onset familial autosomal dominant AD families. Thirty missense mutations have now been described in 52 early onset families world wide in the PS-1 gene. These mutations may account for approximately 70-80% of all early onset autosomal dominant AD families. Only two mutations have been described in the PS-2 gene. These mutations account for 1-2% of all early onset autosomal dominant AD families. Studies into the genetics of late onset Alzheimer's disease have not yet identified a disease-causing mutation. However, one gene has been identified as a risk-enhancing modifier. The apolipoprotein E gene, and more specifically the e4 allele, has been estimated to play a role in the pathogenesis of approximately 50% of all late onset AD cases and has been shown to be associated with AD cases from North America, South America, Asia, Australia and Europe. 1.4.1 B-Amyloid The accumulation of senile plaques in brain parenchyma is a defining neuropathological feature of Alzheimer's disease. These senile plaques are mainly composed of B-amyloid protein. The B-amyloid protein (BA4) is a 4 kilodalton proteolysis product of a larger precursor protein, the B-amyloid precursor protein (BAPP). The BAPP gene is localized to chromosome 21q21.2 and is encoded by 19 exons, of which exons 16 and 17 encode in part for the BA4. Three major splicing variants have been identified containing the B-amyloid sequence, of which APP695 is the major isoform found in the brain Ponte et al., 1988). The two other common isorforms, APP751 and APP770, include exon 7 and exons 7 and 8, respectively (Ponte et al., 1988). 12 1.4.1.1 BAmyloid Protein Proteolysis Pathways The p-amyloid precursor protein comprises a complex group of membrane anchored polypeptides, 100-400 KDa in relative molecular mass. The pAPP is processed through different proteolytic pathways (Haass et al., 1992(a)) (Reviewed by Haass and Selkoe, 1993). The major pathway is the constitutive secretory pathway which involves alpha secretase cleaving the PAPP between Lys 687 and Leu 688 within the extracellular portion of the pAPP peptide sequence. This cleavage site is located in the center of the pA4 peptide sequence (Golde et al., 1992). pAPP is subsequently cleaved by gamma secretase producing soluble APP (APPs), a 3KDa fragment (P3), which constitutes the amino portion of PA4 , and a membrane bound 7KDa fragment (P7) containing only the carboxyl portion of the pamyloid peptide sequence. This pathway is termed "non-amyloidogenic", as it does not produce intact PA4 and therefore cannot contribute to AD pathology. A second proposed proteolytic pathway for the PAPP involves the reinternalization of the full length membrane-bound pAPP from the cell surface which is then routed into lysosomes where it is degraded into potentially amyloidogenic fragments (Haass et al., 1992 (a)). PA4 protein is processed from pAPP by beta and gamma secretases (see Figure 1.2). Beta secretase cleaves pAPP between Met 671 and Asp 672, at the amino terminus of pAPP, releasing truncated APPs (APPsP) in the medium (Seubert et al., 1993). The remaining 11kda portion, which is potentially amyloidogenic, is retained in the membrane and can then be further processed by gamma secretase into PA4. There is evidence that beta secretase cleavage occurs either on the cell surface or within the acidic compartment of late Golgi or transport vesicles derived from a reinternalization pathway carrying pAPP or fragments thereof (Seubert et al 1993; Haas et al., 1993). 13 Figure 1.2: B Amyloid Precursor Protein1 BA4 APP 693 . Beta Secretase Cleavage Site . A P P 670/671 Alpha Secretase Cleavage Site . APP 692 .Gamma Secretase Cleavage Site APP 717 Transmembrane Domain Neither cellular injury nor aberrant proteolysis of BAPP is required for BA4 generation in cells, and it is generated continuously as a soluble peptide during normal cellular metabolism. Intact soluble BA4 protein is released in the medium of normal neuronal cells and non-neuronal cultured cells (Haass et al., 1992 (b)). It is also found in cerebrospinal fluid of non-demented individuals, which might explain why diffuse plaques are commonly seen in 28% of brains of non-demented individuals aged 60 years (Seubert et al., 1992; Gedye et al., 1989). a Figure adapted from Hendriks 1996 14 1.4.1.2 BAPP and Early Onset Familial Autosomal Dominant A D The BAPP gene was first suggested as a candidate gene when researchers characterized its protein as the major constituent protein of senile plaques found in both AD and Down syndrome patients (Masters et al., 1985). The gene mapped to chromosome 21, within the specific portion required in trisomic condition for Down syndrome to result. The finding that an extra copy of the BAPP gene invariably leads to typical AD neuropathology that begins with amyloid plaque formation was important evidence that the metabolism of pAPP into pA4 may play a crucial role in the pathogenesis of AD (Haas et al., 1993). Subsequently, mutations in the PAPP gene were characterized and found to segregate in early onset familial autosomal dominant AD families. To date, six PAPP mutations have been found to segregate in early onset autosomal dominant AD families. Each of these mutations is found within exons 16 and 17 and constitutes a single amino acid substitution. To date, observed substitutions are as follows: three different substitutions at position 717 (Goate et al., 1991; Chartier-Harlin et al 1991; Murrell et al 1991), one at position 692 (Hendriks et al., 1992), one at position 693 (Hardy et al., 1992), and a double mutation at positions 670/671 (Mullan et al., 1992). 1.4.1.3 The A P P 717 Mutations The APP 717 valine to isoleucine substitution was the first mutation ever to be characterized in AD (Goate et al., 1991). Two more pAPP point mutations have now been characterized, a valine to glycine substitution (Chartier-Harlin et al., 1991) and one to phenylalanine (Murrell et al., 1991). The valine to isoleucine substitution appears to be the most common PAPP mutation and has been found in families of different ethnicity, e.g. 15 North American, European, Asian (Goate et al., 1991; Yoshioka et al., 1991). The prevalence of the BAPP717 valine to isoleucine mutation is 5.5% among familial autosomal dominant early onset AD cases (Van Broeckhoven 1995). BAPP717 mutations have been shown to increase the ratio of BA 1-42(43) to PA ^ by as much as two fold (Maruyama et al., 1996; Suzuki et al., 1994), while maintaining the total amount of secreted p-amyloid peptide (Cai et al., 1993; Felsensetein et al., 1994). This mutation was initially thought to be pathologic because longer p-amyloid peptides tend to aggregate faster since the nucleation step is significantly accelerated (Jarrette & Lansbury, 1993). However, it has recently been shown that DNA fragmentation is induced in neuronal cells expressing the PAPP 717 mutants (Yamatsuji et al., 1996). Transfection of neuronal cells with wild type pAPP and with the PAPP 717 mutations showed DNA fragmentation to be increased 4-fold in the cells carrying the mutations compared to the cells carrying the wild type DNA. Nucleosomal DNA fragmentation is one criterion for programmed cell death, suggesting that PAPP 717 mutants can cause apoptosis (cell death) when expressed in neuronal cells. It was shown that the increased PA 1-42(43) production is not the cause of heightened nucleosomal DNA fragmentation. Rather, the cytoplasmic domain (His657-Lys676), normally the ligand-dependent domain of APP695 responsible for the constitutive activation of G 0 (Nishimoto et al., 1993), is critical for the neurotoxicity of the mutant PAPP 717 proteins. This region of the PAPP protein has already been shown to bind G 0 protein. Therefore, it is possible that a pAPP mutation(s) which is close to the Go-activating domain may perturb the normal coupling of PAPP to a G0-mediated signaling pathway by causing unregulated activation of G 0 and, by inference, of a cellular signaling pathway downstream of G 0 . Therefore, it is now believed that G 0 mediates the DNA fragmentation 16 caused by the PAPP 717 mutants and PA4 itself does not appear to play a causative role in inducing DNA fragmentation. Apoptosis as a mode of neurodegeneration in AD is supported by the findings of Su et al. (1994) and Smale et al. (1995) who noted that apoptosis is increased in AD postmortem brains compared to control brains. 1.4.1.4 The Swedish Double Mutation- A P P 670/671 The Swedish mutation consists of a double mutation in the form of a lysine to asparagine substitution at position 670 and a methionine to leucine substitution at position 671 of the pAPP, situated near the amino terminal beta-secretase cleavage site, within the extracellular part of PAPP (Mullan et al., 1992). Transfection of this mutation in both neuronal and non-neuronal cells causes a 3- to 8-fold increase of pA4 production and a concomitant decrease in the production of P3, the 3KDa carboxy terminal subpeptide of PA4 which results from both the alpha and gamma secretase activities (Citron et al., 1992; Cai et al., 1993). The substitution at position 671 has been shown to cause the PA4 increase. However, it is hypothesized that when this is coupled with the substitution at position 670, cleavage may be enhanced (Citron et al., 1992). Given the location of the Swedish double mutation, it is hypothesized to increase the cleavage rate at the beta-secretase site in relation to the alpha-secretase, with a subsequent gamma-secretase cleavage to give PA4 (Felsenstein et al., 1994). The average age of onset for individuals from two Swedish pedigrees with this double mutation is 55 years. It is generally thought that the increase in PA4 associated with this double mutation accelerates the pA4 aggregation and thus the formation of senile plaques. PA 1-40 remains the predominant species in these individual, but a significant amount of pA 1-42(43), the predominant species in senile plaques, is also present (Hendriks et al., 1996)(Dovey et 17 al., 1993). The fact that the increase in BA4 production is detected in AD patients with the double mutation, as well as in presymptomatic persons carrying the Swedish mutation, points to a causal rather than secondary effect of the BA4 overproduction (Citron et al., 1994). 1.4.1.5 The Flemish and Dutch Mutations A mutation consisting of an alanine to glycine substitution at position 21 of the BA4, situated close to the alpha-secretase cleavage site, was found in one Dutch four generation pedigree (Hendriks et al., 1992) hereafter referred to as the "Dutch mutation". This mutation, when transfected into kidney cells, doubles the ratio of BA4 in relation to the P3 fragment. This is thought to be mediated by destabilization of the alpha-helical structure near the alpha-secretase cleavage site, which alters the normal processing of BAPP by the alpha secretase (Haass et al., 1994). The "Flemish" mutation, a glutamate to glycine substitution at position 22 of the PA4, was found in one woman diagnosed with early onset AD (Hardy et al., 1992). This mutation has been shown to act by accelerating amyloid fibril formation and aggregation into amyloid plaques (Clements et al., 1993). 1.4.1.6 Possible Mechanisms of Pathogenesis Although neither the exact mechanism of neurodegeneration nor the role of P-amyloid in this cascade is yet understood, recent studies have now given us an insight into these areas. PA4 has been shown to induce oxidative stress and disrupt cellular ion homeostasis, both of which are implicated in the neurodegenerative process of AD. The peptide also activates microglia, the resident immune effector cells in the brain. Microglia 18 are immobilized in AD and may play a central role in the inflammatory process associated with amyloid plaques. Recent studies have shown that BA4 interacts with a cell surface receptor "RAGE" (Receptor for Advanced Glycation End products) to mediate cell adhesion to BA4 and the induction of oxidative stress in microglia, producing reactive oxygen species and excitotoxins (Yan et al., 1996). Along the same lines, it has also been shown that a class A scavenger receptor (SR) mediates adhesion of microglia to BA4 fibrils, leading to secretion of reactive oxygen species and cell immobilization (Khoury et al., 1996). More studies are needed to clarify the exact role that these proteins have in AD pathogenesis. 1.4.2 The Presenilin Genes: PS-1 and PS-2 Two closely related genes, the presenilins, were recently identified as being involved in some early onset (<65 years) autosomal dominant AD families. The first gene, PS-1, located on chromosome 14 at 14q24.3, was identified by positional cloning (Sherrington et al., 1995). The second gene, PS-2, was later localized to chromosome 1 at 1q31-q42, a region previously identified as a candidate region, by expressed sequence tag homology to PS-1 (Levy -Lahad et al., 1995). These two genes show 67% homology, with the transmembrane domains showing the greatest similarities (see figure 1.3). 19 Figure 1.3: Structure of the Presenilin Proteins3 PS-2 Sequences that diverge from PS-1 ^ ^ ^ ^ ^ PS-1 Sequences that diverge from PS-2 ,rjjjjr Site of mutation Both the PS-1 and the PS-2 proteins are encoded by ten exons and are predicted to be integral membrane proteins with seven membrane spanning domains and a large exposed hydrophilic loop domain between the sixth and seventh transmembrane domains, which makes them appear to have the topology of a receptor, a channel protein or a structural membrane polypeptide (Sherrington et al., 1995). These observations imply that these genes are members of a family of genes (presenilins) with related functions. The exact functions of these genes are not yet known. Nevertheless, a few clues have been provided by their homology to two Caenorhabditis elegans proteins, Sel-12, with which they share 45% homology, and Spe-4, with which they share 20% homology. Spe-4 in C. elegans is a membrane protein involved in intracellular transport or interactions with a Figure adapted from Barinaga 1995 20 fibrillar proteins, processes that are relevant in cytopathological mechanisms occurring in AD (Deng et al., 1996). This has led to the hypothesis that the presenilins may participate in the intracellular BAPP transport or processing. The C. elegans Sel-12 protein functions in receiving cells to facilitate (i) intercellular signaling, (ii) localization and (iii) recycling of membrane spanning molecules, mediated by lin-12 and g1p-1, members of the lin-12/Notch family, which are receptors to specific cell fate (Kamino et al., 1996). Thus, it is conceivable that mutations in the PS-1 and PS-2 genes could alter intraneuronal trafficking of BAPP, causing it to be transported to an intracellular compartment that is more condusive to the generation of the BA4 peptide. To date, 30 different missense mutations in the PS-1 gene have been reported in 52 unrelated early onset autosomal dominant AD families of different ethnic background (See Cruts et al., 1996 for review). Each mutation involves amino acids that are conserved between the PS-1 and PS-2 genes (see Figure 1.3). The mutations are scattered over the PS-1 protein with most mutations located within or just outside one of the seven transmembrane domains. The location of these mutations suggests that they may interfere either with the alpha-helical structure of the transmembrane domains (Alzheimer Collaborative Group 1995) or by aberrantly affecting the anchoring of the protein in the membrane (Wasco et al., 1995). To date, mutations have been found in seven of the ten exons encoding the protein, with mutations in exons 5, 8, 9 and 11 accounting for 75% of all the mutations. The majority of these mutations occur in exon 5 (30%), which encodes the transmembrane domain II, or in exon 8 (34%), which encodes the beginning of the sixth hydrophilic loop. Only two pathogenic mutations for AD have been found in the PS-2 gene on chromosome 1. A 239 methionine to valine substitution was identified in an extended pedigree of Italian origin (Rogaev et al., 1995). The other mutation, a 141 asparagine to 21 isoleucine substitution, was found in affected individuals from seven extended pedigrees of Volga German ancestry (Rogaev et al., 1995; Levy-Lahad et al., 1995), possibly reflecting a founder effect. 1.4.2.1 Expression Patterns PS-1 and PS-2 are expressed in most human tissues, e.g. in heart, liver, pancreas, spleen, kidney, testis and brain (Sherrington et al., 1995; Levy-Lahad et al., 1995; Suzuki et al., 1996). In rat, mouse and human brain tissues, PS-1 and PS-2 transcripts are present in the neuronal cell populations of the hippocampus, cerebral cortex, cerebellum and particularly in the choroid plexus (Kovacs et al., 1996; Suzuki et al., 1996; Moussaoui et al., 1996; Boissiere et al., 1996). Expression is essentially similar in the brains of controls and sporadic AD cases (Nishiyama et al., 1996). PS-1 mRNA is also expressed at low levels in glial cells (Suzuki et al., 1996). Immunohistochemical analysis of brain sections of Alzheimer's disease patients has shown that PS-1 is localized to senile plaques in a similar pattern to p-amyloid peptide (Wisniewski et al., 1995). However, PS-1 mRNA expression levels do not vary significantly between AD subjects and controls (Johnston et al., 1996). Immunocytochemical analysis of cultured cells have also localized presenilin proteins to the perinuclear zone and intracellular structures of the cytosol. PS-1 seems to have a vesicular sublocalization (Moussaoui et al., 1996; Kovacs et al., 1996) and is associated with rough endoplasmic reticulum membrane, perinuclear membrane and plasma membrane in cells (Takashima et al., 1996; Cook et al., 1996). This localization suggests that PS-1 is an integral membrane protein, as predicted by its amino acid sequence. 22 1.4.2.2 Age of Onset Patients with PS-1 mutations usually have an age of onset between 35 and 50 years. The age of onset is determined by both the position and the nature of the mutation within the PS-1 gene, as unrelated families with the same mutations have a very similar age of onset. This leaves little room for other genetic or non-genetic (environmental) modifying factors (Kamino et al., 1996). To date, the one exception is a 68 year old individual carrying a PS-1 mutation who is cognitively normal, from a family with an average age of onset of dementia of 55 years (Rossor et al., 1996). Apolipoprotein E genotype does not seem to affect the age of onset in individuals carrying PS-1 and PS-2 mutations (Sorbi et al., 1995; Sherrington et al., 1996; Poduslo et al., 1996). Individuals with a PS-2 mutation tend to have a later age of onset than do PS-1 carriers, and the range of onset age is highly variable between early onset PS-2 families. Some individuals have onset in their forties, while others only have onset in their seventies (Levy-Lahad et al., 1995). The variability, however, is not affected by the apoE genotype, suggesting that other genetic and/or non-genetic (environmental) factors modify the age of onset for carriers of a PS-2 mutation (Sherrington et al., 1996). 1.4.2.3 Possible Functions of the Presenilins Although the function(s) of the presenilins are not yet known, several possibilities have now been suggested based on their location in the cell. These include a function in neuritic transport (Moussaoui et al., 1996) or as a cell adhesion molecule (Takashima et al., 1996). There is now evidence that PS-1 and PS-2 mutations cause, or at least promote, an early and excessive deposition of BA 1.42 within the brain (Mann et al., 1996; Sheuner et al., 1996). Similar to AD patients carrying the APP 717 mutation, AD patients with a PS-1 or 23 PS-2 mutation have a higher amount of plasma PA ^ 2 (Lemere et al., 1996), as do asymptomatic carriers of PS-1 or PS-2 mutations, compared to sporadic AD patients. These observations support the hypothesis that the presenilis have a role in the processing of PAPP. The early deposition of PA into senile plaques in PS-1 patients also lends credence to PA being a cause of AD and not simply an effect (Iwatsubo et al., 1996). Presenilin missense mutations may also alter membrane protein trafficking, enhancing the exposure of pAPP to the gamma-secretase that cleaves at PA position 42, and thereby increasing PA ^ 2 generation (Lemere et al., 1996). 1.4.2.4 PS-1 and Late Onset Alzheimer's Disease Given that the age of AD onset in PS-1 carrying individuals is strongly correlated with the position and nature of the mutation, a role for PS-1 mutations in late onset AD may exist because (i) allele sharing between affected family members with late-onset (ages 60 -70) AD, has been observed (Shellenberg et al., 1993), and (2) at least four identified PS-1 mutations have been linked to families with AD onset in the late fifties (Campion et al., 1995) . A recent study identified a common polymorphism of the PS-1 gene within the intron 3' to exon 8 region where homozygosity of the "1" allele was associated with a doubling of the risk for late-onset AD, compared to 1/2 heterozygotes and 2/2 homozygotes (Wragg et al., 1996). How this intronic polymorphism increases the risk of AD is unclear, but a possibility is that the polymorphism is in linkage disequilibrium with another mutation found in the PS-1 gene. Several studies have tried to reproduce these results, with conflicting findings (Higuchi et al., 1996; Kehoe et al., 1996; Scott et al., 1996 (a & b); Perez-Tur et al., 1996) . 2 4 1.4.3 Apolipoprotein E The mature form of Apolipoprotein E (ApoE) present in human plasma and cerebrospinal fluid is a single glycosylated 37 kilodalton polypeptide containing 299 amino acids. The Apo E gene has three common forms (e2, e3, e4) accounting for 99% of all Apo E isoforms. These three alleles are distinguished by a simple amino acid substitution at residues 112 and 158 (see Table 1.1). Table 1.1 : Apolipoprotein E polymorphisms Allele Position 112 Position 158 e2 cystein cystein £3 cystein arginine £4 arginine arginine These three alleles give rise to three homozygote phenotypes: e2/£2, E3/E3, E4/E4, and to three heterozygote phenotypes: E2/E3, E3/E4, and E2/E4. In European and North American Caucasian populations, the E3 allele is the most common, with an allele frequency of 0.77 (Walden & Hegele, 1994). Apolipoprotein E is a major constituent of several serum lipoprotein particles, including (i) chylomicron remnants, which transport dietary triglyceride and cholesterol from the intestine to the liver and peripheral tissues, (ii) very low density lipoproteins (VLDL), which function primarily in the transport of triglycerides from the liver to peripheral tissues, and (iii) a subclass of high density lipoproteins (HDL), which function in cholesterol redistribution among cells (Mahley 1988). Apo E on these particles serves as the ligand 25 responsible for recognition by two cell surface receptors, a hepatocyte receptor specific for Apo E and a receptor on many cell types, the Apo B,E (LDL receptor) receptor, that recognizes Apo E as well as another apolipoprotein, apo B. Binding of ApoE and Apo B-containing lipoproteins to these receptors stimulates receptor-mediated uptake of these particles, which serve as a primary source of cholesterol for normal cellular metabolism (Ignatius et al., 1986). Apo E is unique among the apolipoproteins because of its special relevance to nervous tissue. Unlike other apolipoproteins that are primarily produced in the liver and small intestine (Walden & Hegele, 1994), ApoE is also produced substantially by Schwann cells in the Peripheral Nervous System (PNS) and by astrocytes and oligodendrocytes in the Central Nervous System (CNS) and comprises the major apolipoprotein in the cerebrospinal fluid (Namba et al., 1991). ApoE coordinates the mobilization and redistribution of cholesterol in repair, growth and maintenance of myelin and neuronal membranes during development or after injury in the PNS (Mahley 1988; Ignatius et al., 1986; Namba et al., 1991; Muller et al., 1985; Skene & Shooter 1983). Secreted ApoE accounts for up to 2-5% of the total soluble protein in the extracellular space surrounding the injured neuronal stump (Skene & Shooter, 1983). In CNS, ApoE has a pivotal role in the mobilization and redistribution of cholesterol and phospholipids during membrane remodeling associated with synaptic plasticity (Mahley 1988). 1.4.3.1 A p o E and Alzheimer's Disease In recent years, many studies have been focused on the putative association between the e4 allele of the Apolipoprotein E and AD (e.g. Brousseau et al., 1994; Kuusisto et al., 1994; Peacock & Fink 1994;Saunders et al., 1993b; Martins et al., 1995; Hong et al., 1996). The frequency of e4 is greatly increased in late onset sporadic AD cases (Saunders et al., 1993a; Poirier et al., 1993; Nalbantoglu et al., 1994), late onset familial AD cases (Strittmatter et al., 1993; Corder et al., 1993), early onset sporadic AD cases (Okuizumi et 26 al., 1994; Mayeux et al., 1993), and in early onset familial AD cases (Lannfelt et al., 1994; Van Duijn et al., 1994) compared to the general population. The E4 allele frequency in AD populations ranges from 0.38-0.50 compared with 0.12 - 0.16 in the general population (Strittmatter et al., 1993; Poirier et al., 1993). This association has been reported for both early and late onset AD cases, with the exception of early onset families linked to chromosome 14 and chromosome 1 (Van Broeckhoven et al., 1994; Sherrington et al., 1996). Approximately 75% of the AD patients with an age of onset of 55 to 60 possess at least one E4 allele, but this percentage drops to approximately 30% of the patients with an age of onset of >90 (Rebeck et al., 1994; Sobel et al., 1995; Payami et al., 1995; Asada et al., 1996). The decrease of the E4 allele frequency in the very old could be attributed to poorer survival because of cardiovascular disease or earlier dementia. A dosage effect for the E4 allele of Apo E has been reported. A correlation has been reported in late onset AD families, between age of AD onset and the number of copies of the E4 allele, with homozygotes (E4/E4) having an earlier age of onset. The presence of each additional copy of the E4 allele seems to increase the risk to develop AD and decrease the age of AD onset (Corder et al., 1993, Poirier et al., 1993). Table 1.2 presents the results of a study on 42 families with late onset AD (Corder et al., 1993). As can be seen, the average age of onset decreases by approximately 8.5 years with each additional copy of the E4 allele. Similar dosage effects of the E4 allele on the risk for AD and the age of AD onset have been detected in late onset sporadic AD (Mayeux et al., 1993; Poirier et al., 1993; Nalbantoglu et al., 1994). 27 Table 1.2: The Effects of ApoE Genotypes on Risk and Age of AD Onset Apo E Genotype Average Age of Onset Proportion Affected (%) (years) e2/e3 or e3/e3 85 20 E4/EX 76 47 E4/E4 68 91 Despite the above, the presence of an E4 allele is not essential for AD (Corder et al., 1993). Although Apo E E4 enhances the risk of AD, AD can clearly occur in the absence of an E4 allele (Corder et al., 1993; Roses et al., 1994). Conversely, the presence of one or two E4 allele(s) does not necessarily result in the development of AD (Sobel et al., 1995; Rebeck et al., 1994; West et al., 1994). The relationship between the E2 allele of Apo E and AD is unclear. A protective effect of the Apo E E2 allele against the development of AD has been reported (Corder et al., 1994; Benjamin etal., 1994; Smith et al., 1994; Talbot et al., 1994). AD patients carrying an E2 allele seem to have a lesser degree of neuropathological changes, such as cortical neurofibrillary tangles and cortical senile plaques, in comparison with AD patients without a copy of the E2 allele (Benjamin et al., 1994). In an autosomal dominant early onset AD family where a Val->lle BAPP717 substitution was identified, the presence of an E2 allele in an individual with the BAPP mutation appeared to delay the onset age for AD (Alzheimer's Disease Collaborative Group 1993). However, this observation is not universal. A study on the African-American population showed that individuals with an E2 allele actually 28 had an 8-fold increased risk for AD compared with e3/e3 individuals (Maestre et al., 1995). This finding, however, was not found for either the Hispanic population or the Caucasian population. Recently, Hendrie et al. (1995) did not find evidence for a protective role for the e2 allele in a community study of elderly African-Americans. The observed association between the e4 allele of ApoE and AD could result from a direct involvement of the ApoE transport system in the pathophysiological processes contributing to AD. Conversely the ApoE e4 allele may be in linkage disequilibrium with another gene(s) responsible for the increased risk to AD. However, abundant support for the direct involvement of ApoE in the pathophysiological processes contributing to AD has been demonstrated. Several studies have now shown that late onset AD patients homozygous for the E4 allele have increased vascular and plaque amyloid deposits compared with E3/E3 homozygotes or E3/E4 heterozygotes (Polvikoski et al., 1995; Nagy et al., 1995; Schmechel et al., 1993). A statistically significant association has been reported between the degree of BA deposition in both vessels and plaques and the dosage of ApoE e4 alleles. The demonstration that ApoE binds BA4 in vitro (Strittmatter et al., 1993) and amyloid plaques in vivo (Namba et al., 1991; Rebeck et al., 1993) along with the observation that LRP (LDL receptor-related protein), which is capable of binding ApoE-enriched lipoprotein molecules, is associated with senile plaques provides further evidence to support the hypothesis of a direct interaction between BA4 and ApoE in the development of sporadic late onset AD. Specifically, the e4 allele of ApoE was found to have a greater binding affinity to synthetic amyloid B peptide than did the E3 allele. ApoE, and in particular the E4 allele, has also been shown to accelerate amyloid fibril formation, which precludes the formation of senile plaques (Wisniewski et al., 1994). 29 ApoE has also been found to bind to intracellular neurofibrillary tangles in vivo (Namba et al., 1991). It has been hypothesized that the e4 allele of ApoE does not bind to tau protein, therefore not altering the rate of tau-phosphorylation and self-assembly into paired helical filaments (PHFS) (Strittmatter et al., 1994). Tau protein promotes microtubule assembly and stabilizes microtubules. Tau's ability to bind to microtubules is partly determined by the number of attached phosphate groups. Hyperphosphorylated tau does not bind, thereby destabilizing microtubules. Over time, a bias toward destabilization of microtubules and the formation of neurofibrillary tangles may occur in individuals who inherit ApoE s4 alleles, leading to a shorter functional neuronal life span. An increase in the number of NFTs in e4 carrying individuals in comparison to E3/E3 homozygdtes has been found (Polvikoski et al., 1995; Nagy et al., 1995), although some have failed to find such a relationship (Landen et al., 1996). Another theory in the formation of neurofibrillary tangles involves the suppression of phosphatases, enzymes that remove phosphate groups from tau, in neurons of AD patients (Matsuo et al., 1994). In one study on biopsy brain tissue, it was discovered that tau proteins of non demented individuals were equally phosphorylated as PHF-tau, however these tau proteins were rapidly dephosphorylated in the post-surgical period, while the PHF-tau remained hyperphosphorylated (Matsuo et al., 1994). 1.5 Transgenic Models Transgenic animal models of AD could be very useful in the search for the answer to the Alzheimer's puzzle. Several mouse models now exist which express a mutated human APP gene. However, the search for an appropriate animal model has been difficult. Models demonstrating behavioral problems did not have senile plaques; models with senile plaques did not display the expected behavioral problems (Games et al., 1995). Recently, a transgenic mouse strain expressing the human Swedish APP double mutation also expressed the behavioral, biochemical and pathological abnormalities associated with 30 human AD (Hsiao et al., 1996). This animal model now makes it possible to test drugs that inhibit amyloid deposition into plaques. If these drugs show inhibition of amyloid plaque formation and memory impairments in the transgenic mice, it would be the best evidence yet that the two are causally related and that BA4 deposition is not just a secondary cause of AD. Studies of mice transgenic for the PS-1 mutation have recently shown that PS-1 mutations increase the BA4 1 . 4 2 concentration in brain (Duff et al., 1996), thus supporting the hypothesized role of PS-1 mutations in humans. 1.6 Empiric Risk Estimate Studies in Alzheimer's Disease Genetic epidemiological studies of familial aggregation can test whether the disease under study, e.g. AD, occurs more often among biological relatives of affected individuals than in the general population. Data from these studies can be used for genetic modelling. Although numerous studies have now addressed the familial aggregation in AD, study designs have varied. Early studies showed a lifetime cumulative risk approaching 50% for first degree relatives of AD index cases who survived well into the ninth decade of life (Breitner et al., 1988; Huff et al., 1988). These results were taken as support for an autosomal dominant mode of inheritance for all AD cases. However, these studies were based on relatively small sample sizes, especially in the eighth and ninth decades of life, which leads to unstable risk estimates. More recent studies have consistently shown a cumulative lifetime risk between 23-30% for first degree relatives of AD cases, compared to 10% for first degree relatives of controls (Sadovnick et al., 1989; Farrer et al., 1989; Hirst et al., 1994; Silverman et al., 1994 (a)8i(b)). These studies have shown that familial factors do have an impact on the risk of AD (Silverman et al., 1994(a) &(b); Li et al., 1995), as first degree 31 relatives of AD cases are at an increased risk of AD compared to first degree relatives of controls, and the age of onset in first degree relatives of AD cases is earlier than in first degree relatives of controls. Although familial factors have an effect on the risk of AD even in first degree relatives of late onset cases (Li et al., 1995), the impact of these factors may begin to decline in the seventh or eighth decade of life relative to other causes, such as advancing age (Silverman et al., 1994(a)). Gender and relation (sib or parent of index case) have been tested as independent covariates in some studies. Although relation does not seem to modify the risk to first degree relatives (Silverman et al., 1994 (b); Farrer et al., 1989), the effect of gender remains controversial (Silverman et al., 1994 (b); Li et al., 1995; Farrer et al., 1989). 1.7 Non-Autosomal Dominant Alzheimer's Disease Although studies into the influence of genetic factors specifically in non-autosomal dominant AD cases have not been done, there are indications that genetic influences do exist in these strictly non-autosomal dominant AD cases. As mentioned in section 1.4.3.1, apolipoprotein E e4 allele is associated with an increased risk of AD in both early and late onset AD cases and decreases the age of onset of AD in a dose dependent manner in familial and sporadic late onset AD cases. Four major twin studies have now shown evidence for genetic factors in AD (Breitner et al., 1995; Bergem 1994; Johansson et al., 1992; Raiha et al., 1996). Although these studies were not limited to strictly non-autosomal dominant AD cases, the concordance rates of 21-83% for monozygotic twin pairs (MZ) and 9-42% for dizygotic twin pairs (DZ) can be taken as an indication of genetic factors in non-autosomal dominant AD, as these cases represent 90% of all AD cases (Cruts 1996). AD concordance rates in twin studies depend heavily on the age of the twin sample (Breitner et al., 1993), which explains the lower 32 concordance rates seen in the US study (21% for MZ and 11% for DZ), where the twin pairs tended to be of younger age. Adoption studies have not yet been attempted in AD (to the best of my knowledge). Because AD is a late onset disease, adoption studies would be very difficult and would require very long follow-up periods. Age-specific incidence rates of AD have been shown to be similar across populations (USA, Japan, France, Sweden, UK) up to age 75 (Van Duijn 1996). These results indicate that the genetic component associated with AD could be a common polymorphism that is present in several populations. A common polymorphism would also agree with the finding that there is no difference in the prevalence of AD in three ethnic groups (African-Americans, Hispanics, and Caucasians) living in the same community (Maestre et al., 1995). There is also evidence for protective genetic factors in non-autosomal dominant AD, as shown by a recent study in which the relationship between the genetic degree of Cherokee ancestry and the development of AD was looked at in individuals from the Cherokee Nation (Rosenberg et al., 1996). Results from the study showed that the risk of AD increased with decreasing degree of Cherokee ancestry and the risk of AD increased 9-fold with every 10% of Cherokee ancestry lost. This study was relatively small (n=26), and therefore more studies are needed to confirm these findings (Rosenberg et al., 1996) 33 1.8 Clinic for Alzheimer Disease and Related Disorders, University Hospital-UBC Site The Clinic for Alzheimer Disease and Related Disorders, University Hospital- UBC Site (AD Clinic) has the following mandate: i) to provide assessment, often as a second opinion, for individuals referred with memory problems and other cognitive impairments; ii) to provide counseling to patients and their family members with respect to prognosis and familial risks; and iii) to conduct research on Alzheimer's disease and related disorders. Patients attending the AD Clinic are assessed by a multidisciplinary team composed of an internist/geriatrician, a neurologist, a neuropsychologist, a social worker, a geneticist, and a speech-language pathologist. Each patient undergoes a series of tests including blood work, CT head scan, chest x-rays, electrocardiogram, and a wide series of neuropsychological tests in addition to clinical exam(s). Once the results of all these assessments are available and reviewed, a consensus diagnosis is assigned according to the NINCDS-ADRDA criteria (McKhann et al., 1984). Patients meeting diagnostic criteria for a dementing illness are followed longitudinally whenever possible. Any individual who, at initial assessment, does not fulfill the criteria for dementia is followed and re-assessed at regular intervals. Any change in his/her diagnosis is noted. 1.9 The Canadian Study on Health and Aging The Canadian Study on Health and Aging (CSHA) is a longitudinal population-based study of the elderly Canadian population. This study was designed in response to the need for standard screening and diagnostic criteria for dementia, including AD, across the nation. The objectives of the CSHA were; i) to estimate the prevalence of dementia among elderly Canadians with the use of a common research protocol; ii) to determine the risk factors for 34 Alzheimer's disease; iii) to describe the current patterns of caring for patients with dementia in Canada and to assess the burden on caregivers and their need for support; iv) to establish a uniform database for studying the natural history of dementia as well as planning and evaluating interventions. CSHA subjects were randomly drawn from a representative sample of the elderly Canadian population (aged 65 years and over) living in the community and in institutions. Guidelines for the sample selection were as follows: (A) Community Sample 1. Select individuals aged 65 and over as of October 1, 1990, from provincial health registry. 2. Exclude individuals living in institutions. 3. Use postal codes to select individuals living in designated study areas. 4. Create sequential sampling frames alphabetically for each Central Metropolitan area and Urban Area using postal codes. 5. Sort the sample into the Central Metropolitan Areas and Urban Areas using postal codes. 6. Sort the above samples into three age strata : 65-74, 75-84,85+ 7. Randomly select the number (2N) subjects required for each stratum. By selecting twice the required number per stratum, replacement cases are available. 8. Within each stratum, the odd numbers constitute the sample, and the even numbers, the replacement sample. 35 (B) Institutionalized Sample 1. Use the BC Central Registry of Continuing Care Facilities to obtain a list of all institutions within Vancouver, Victoria, Kelowna, Kamloops, Matsqui, Mission, Chilliwack and Nanaimo. Geographic location was identified by postal codes. 2. Obtain a list of all individuals aged 65 and over as of Oct. 1, 1990. 3. Sort the list into three strata based on size (maximum number of beds) of institution: small (4 - 25 beds), medium (26 -100 beds) and large (>100 beds). 4. From each stratum, collect a random sample of approximately 200 (to allow for a final sample of 84 individuals per stratum with replacements). The Modified Mini-Mental State (3MS) Examination was selected as the screening procedure for cognitive impairment (CSHA 1994 (a)). The 3MS assesses a subject's cognitive functioning, examining his/her orientation to time and place, instant recall, and short term memory (Teng & Chui 1987). The test is scored by summing the points assigned to each task. A score of 77 or less out of 100 generally indicates some form of cognitive impairment. The lower the score, the greater the likelihood of cognitive impairment. For the purpose of this study, only individuals scoring 80 or above on the 3MS test were "controls" i.e. free of cognitive loss or dementia. 36 2. Objectives of the Study The primary objective is to test the hypothesis that early onset (<65 years) non-autosomal dominant AD is more genetically loaded than late onset (>65 years) non-autosomal dominant AD. Secondary aims are : (i) to examine the effect(s) of the gender of the index case on the risk of AD to first degree relatives. (ii) to examine the effect(s) of the gender of the relative on their AD risk. (iii) to examine the effect(s) of the relationship between the index case and the first degree relative (i.e. sib or parent) on the relative's risk of AD. (iv) to compare the risk estimates for first degree relatives of non-autosomal dominant AD cases with those for first degree relatives of controls to determine genetic loading. (v) if certain subgroups of first degree relatives of AD cases are indeed at an increased risk, to determine whether this simply reflects the general population. 37 3. Relevance of the Study 3.1 The Significance of Alzheimer's Disease in the Canadian Population In 1991,10.6% of the Canadian population was aged 65 years and over. This proportion is expected to rise to 21.8% in 2031 (CSHA 1994 (a)), when it is estimated that 7.8 million Canadians will be aged 65 years and above. This could have major implications on the costs of providing care. In 1991, it was estimated that the net economic cost of dementia in Canada was $3.9 billion, or 5.8% of Canada's total health care cost for that year. These numbers only address the financial cost of direct health care for an AD patient. They do not take into account the immense financial, emotional and physical burden AD places on family members of affected individuals. As mentioned earlier (see figure 1.1), AD accounts for up to two thirds of all dementia cases in the elderly (>65 years) Canadian population. Therefore research into the cause of non-autosomal dominant AD is of great importance. If there is evidence of increased familial aggregation in these families, it will then be possible to search for the genetic and non-genetic factors that have a role in this form of AD. Once the genetic causes of non-autosomal dominant AD are found, it will be possible to study the normal function of affected genes, and thus move closer to effective treatments, prevention and in turn a cure. It is also important to search for non-genetic risk factors, as delaying onset of AD for a few years by simple modification of non-genetic factors could prevent an individual from becoming affected, by pushing onset past the individual's natural death by another cause. 38 3.2 Risk Assessment for Relatives of Alzheimer's Disease Cases Increasing public awareness about AD means that it is no longer a "hidden" disease. With a prevalence of up to 22% in the Canadian population aged 85 and above (CSHA 1994(a)), much of the population has been directly or indirectly touched by this disease. Although, as discussed in section 1.4, specific mutations can be identified in selected early onset autosomal dominant AD families, these families in reality only represent an extremely small proportion of the total AD population (see Figure 3.1). In these rare families, predictive testing can be offered for asymptomatic individuals, although this is not a routine or frivolous undertaking (Sadovnick et al., 1997). Figure 3.1: Proportion of AD Cases Caused by Mutations: Sporadic and Familial Cases Combined3 BAPP po-i Other 89% For the well- documented early- and late-onset autosomal dominant AD families with no identified mutation (also a very small proportion of the total AD population), risk counselling can be given based on an autosomal dominant model. However, for the vast a Data taken from Cruts 1996 39 majority of families with at least one affected member, risk counselling cannot include testing of a known mode of inheritance- i.e. autosomal dominant. For these families, risk counselling must rely on the most up to date empiric risk estimates to determine the risk to first degree relatives. Despite the identification of genetic (e.g. ApoE) and non-genetic (e.g. low level of education, head injury) risk factors, it has been well documented that AD can occur in the absence of these risk factors, and AD is not inevitable in the presence of these risk factors. 40 4 Material and Methods 4.1 Cases "Cases" for this study included all consecutive, unrelated patients attending the UBC Alzheimer Disease Clinic, given a diagnosis of "probable" or "definite" Alzheimer's disease according to the NINCDS-ADRDA criteria (McKhann et al., 1987), during the period from August 1985 to August 1995. The NINCDS-ADRDA criteria for definite AD and probable AD are as follows2: Definite AD: (I) the clinical criteria for probable AD must be met, and, (ii) there must be histopathologic evidence obtained from a brain biopsy or brain autopsy. Probable AD: (i) dementia must be established by clinical examination and documented by the Mini-Mental test, Blessed Dementia Scale, or some similar examination, and confirmed by neuropsychological tests, (ii) there must be deficits in two or more areas of cognition, (iii) there must be a progressive worsening of memory and other cognitive functions, (iv) there must be no disturbance of conciousness (v) onset must be between 40 and 90 years, most often after the age of 65, (vi) there must be absence of systemic disorders or other brain diseases that in and of themselves could account for the progressive deficits in memory and cognition. 2 taken from McKhann et al., 1987 pp 940 41 The present study was designed to include only cases with a diagnosis of probable or definite AD from families for which sufficient data are available to make an autosomal dominant mode of AD inheritance unlikely. Historically, the term FAD or familial Alzheimer's disease has been used to refer to loaded (multiple affected members), multigenerational families. The criteria used to identify FAD families in previous studies (Sadovnick et al., 1989; Hirst 1993) are the following: 1) A detailed family history must be available for at least the index case's generation and the previous (parental) generation. 2) Good clinical documentation of dementia in relatives, preferably from at least two separate sibships within the family, must be available; and there must be no other plausible explanation for the dementia (see section 1.2). 3) Neuropathological documentation of Alzheimer's disease should be available for at least one member of the family, but preferably for two or more. 4) Accurate information on ages of death and/or present ages of relatives must be available so that it is possible to assess the "significance" of being clinically unaffected. These criteria are quite stringent, and thus identify a relatively conservative group as FAD. For this thesis, all families meeting the FAD criteria, along with all families in which an autosomal dominant mode of inheritance could not be ruled out were excluded from further study (see Figure 4.1 for examples). Possible autosomal dominant families were identified using the FAD criteria detailed above in a relaxed manner. 42 Exceptions to the FAD criteria are as follows: 1) A detailed family history for the parental generation was not absolutely required. 2) Families in which the transmitting parent (i.e. the parent from which the disease causing mutation is received) did not survive to the at risk age, but sibs of the transmitting parent who survived into the "at risk" age were affected (i.e. the parent could have become affected if he/she would have lived longer) were excluded, (see figure 4.1) Figure 4.1 Autosomal Dominant Pedigrees FAD Family 68 ^ ^ 6 6 Possible Autosomal Dominant Family 68 / 60 55 6"# 55 Ideally, only patients with neuropathological confirmation should be used for studies of this kind. However, in reality only a very small number of AD Clinic patients actually have neuropathological confirmation after death, despite the best intentions of the families and physicians. It has also been shown that cumulative lifetime risk estimates for first degree 43 relatives of definite AD cases do not differ substantially from the cumulative lifetime risk estimates for first degree relatives of clinically diagnosed probable AD cases (Brierer et al., 1992). 4.1.1 Relatives of Cases Detailed information was collected on family members of each case using the "family history" method (Silverman et al., 1986). This method is a modification of the "family study" method which requires that all family members be clinically assessed. The family study method is not feasible in the study of Alzheimer's disease. As AD is a late onset disorder, many of the family members of interest may already be deceased. Another difficulty with the use of the family study method in AD is that patients have often immigrated or moved, and are therefore geographically separated from their families, making direct assessment of each individual family member very problematic. The family history method relies on knowledgeable informants to provide information on relatives of index cases. While this method typically underestimates the number of affected relatives compared to the family study method, it has been shown that the use of multiple informants reduces this underestimation to a level of non-significance (Andreasen et al., 1977; Silverman et al., 1986). For the present study, informants were preferably sibs and/ or the spouse of the index case, as children tend to have less information on the parental generation. Cases and co-informants were interviewed by the AD Clinic geneticist or genetic counsellor who collected and verified all family history data. When possible, relatives identified as "affected" during the family history interview were assessed at the AD Clinic. If this was not feasible, relevant medical records and/or autopsy records were obtained, with proper consent. These records were then reviewed by the AD Clinic team (geriatrician, 44 neurologist) to determine the most likely diagnosis for each individual. In addition, next of kin were asked to complete a "dementia questionnaire" (Silverman et al., 1986) (See appendix A) for all affected family members who were deceased or unable to undergo a clinical assessment. This questionnaire was used in conjunction with health records to improve the accuracy of diagnosis. For the purpose of this study, first degree relatives were only considered affected after all other causes of dementia (see section 1.2) were excluded. The diagnosis of affected was made only after considering all medical records available, and if there was firm evidence of a progressive and irreversible dementia. In the absence of medical records, the dementia questionnaire and informant interview(s) were carefully considered in assigning "affected" or "unaffected" status. This method of verification has been successfully used in other published studies on this AD Clinic population (Hirst et al., 1994; Sadovnick et al., 1989). 4.2 Controls Controls were BC residents who were identified by CSHA and who scored 80 points or above on the Modified Mini-Mental State Exam (3MS) (CSHA 1994(a)). A high cut off score of 80 points on the 3MS test was selected to exclude the vast majority of individuals showing mild forms of cognitive loss and/or those in the very early stages of dementia. 4.2.1 Relatives of Controls Detailed family history information was collected on first degree relatives of controls in a similar manner to first degree relatives of cases. Reportedly affected relatives were assessed according to the protocol described for relatives of cases (see section 4.1.2). 45 4.3 Kaplan-Meier Age-Specific Risk Estimates Age-specific risk estimates for first degree relatives of cases and controls were calculated using the Kaplan-Meier method, also known as the product-limit method (Kaplan & Meier 1958). This is a non-parametric method that makes no assumptions about the age of onset distribution of the first degree relatives. The quantity estimated by the Kaplan-Meier method is the joint probability of (i) inheriting the genotype and (ii) having the associated latent time to expression being less than a fixed age. This method allows the use of censored data, i.e. all first degree relatives of index cases can be used to estimate the risk, including those who die, or are studied before onset can occur (censored individuals). The Kaplan-Meier estimate is unbiased but has the following restrictions: 1) at least one individual in the cohort must be followed past the age at which the "escape probability" (probability of remaining disease free) is to be evaluated, and 2) it is necessary that the i censoring of individuals not be associated with the beginning of onset (Fisher & Van Belle 1993). The major disadvantage of using the Kaplan-Meier lifetime risks method is the need to assign an age of onset for "affected" relatives, which is often problematic. The exact time of AD onset is often unclear as initial signs and symptoms tend to be mild and evolve slowly along a continuum. We defined age of onset as the point when the first detectable symptoms of cognitive embarrassment became apparent. This definition of age of onset is widely used (Li et al., 1995; Korten et al., 1993). However, other studies have used a more conservative estimate of age of onset as either 1) the point at which the first definite symptoms of frank dementia were apparent, or even 2) the point at which progressive dementia was apparent. The use of these different criteria for age of onset can modify the Kaplan-Meier risk estimates substantially (see Breitner & Magruber-Habib 1989 for review). 46 Assigning an age of onset to affected relatives is often complicated by a lack of sufficient information to do this. For example, autopsy confirmation may be available but there may be no records or informant who can help pinpoint the age of onset. In these cases, a conservative method of assigning the latest possible age as the age of onset was used. This method has previously been used successfully (Sadovnick et al., 1989; Hirst 1993) and seems to be the best method for assigning an age of onset in the occurrence of a missing age, as it is based on each patient individually. Other methods of assigning an age of onset for these individuals are based on estimates taken from the total sample of affected relatives i.e., the average age of onset of all affected relatives, the "best" case scenario (all individuals with an unknown age of onset are assigned the oldest age of onset documented in all other affected first degree relatives) or the "worst" case scenario (all individuals with an unknown age of onset are assigned the earliest age of onset documented in all other affected first degree relatives), (see Hirst 1993 for review) The EGRET statistical package (1993) was used for statistical analyses. The Kaplan-Meier age-specific risk estimates can be defined for a group of individuals by using the following terminology: let M represent the ordered and distinct onset times (ti, t2,....tM) of the N first degree relatives of cases. Suppose that cm represents the number of individuals who have onset at time tm (m=1,2 M) and d m represents the number of individuals who are censored in the interval {tm, tm+1}, either by not showing symptoms of AD at the time of study, or by dying before onset could occur. Let n m be the number of relatives surviving at the beginning of the m t h interval that are at risk, then n m = Xj=mM (Cj+dj). 47 The product-limit survivor function, or empirical survivor function, S(t), is defined as: S(tm)=n j = m nj - q "J where n.j - q / nf represents the fraction of relatives in the interval who do not become affected. The survivor function S(tm) represents the cumulative proportion of individuals remaining AD-free at the last age of estimation. The probability density function F(t) represents the proportion of individuals succumbing to the disease. F(t) can be calculated by : F(t) = 1 - S(t). The variance for S(t) can be estimated using Greenwood's formula (Fisher & VanBelle 1993): r j 2 m = S(tm) 2 Xmj=1 Cj nj (nj-q) Differences between two independently estimated lifetime risk curves can be assessed by the Mantel-Hanzel Log Rank test (Crowley & Breslow, 1984). This test can be defined as follows: let ti < t2 < < tL be the ordered, distinct, onset times from the combined data of the two samples. Let ck be the number of onsets at time tj in sample k, (k= 0, 1), and d i k be the number of censored observations from sample k, in the interval [t, tj+i ]. The number at risk from each sample at tj is n i k = Zj=iL (c]k + d j k). The total number 48 of observations in each group is N k = d0k + Xj=iL ( c ik + d i k). Also define c, = c i 0 + d, nj = n i 0 + nii and N =N0 + Ni for both samples combined. The observations at ti can then be presented in a 2 X 2 contingency table, representing those individuals at risk from each sample and whether they become affected or remain free of Alzheimer's disease, as follows: At time ti Total at risk Total number of Total remaining Sample onsets unaffected Co l"lio " Co fin - cM ni Q ni - c The significance of an assumed constant odds ratio or relative risk can then be assessed from independent 2X2 tables. For example, the observed (O) number of onsets at time t in sample 1 is equal to dM, while the expected number of onsets (E) (if the risk of onset is equal in the two groups) is E = ni! X d/ni. The overall test consists of adding the difference between observed and expected, over tables for each age of onset to obtain 0^ - Ei = I i = 1 L (Oii - En ). The Log Rank test statistic is then ( d - f IV, with the variance, V, defined as: V = Xi=iL riio riji d| (^  - dj) (ni)2 (n,-1) 49 The Log Rank statistic is distributed approximately as a Chi- Square (%2) with one degree of freedom. This statistic compares the two samples in either of the two different directions (see Crowley and Breslow 1984 for review). The difference between the estimated cumulative lifetime risk estimates was calculated by the difference of proportions test: Z = S i ( T M ) - S 2 ( T M ) { S E ( S 1 ( T M ) ) 2 + S E ( S 2 ( T M ) ) 2 } 1 / 2 where S ^ T M ) and S 2(TM) are the estimated cumulative lifetime risks for group 1 and 2 respectively, and SE(S 1 (T M ) ) and S E (S 2(TM)) are the standard errors for both estimates respectively (Fisher & Van Belle 1993). 4.4 Cox Proportional Hazard Ratios The Cox proportional hazards model, used to calculate odds ratios for variables of interest, is a method of analyzing survival times based on the assumption that each variable ( X L X 2 , X P ) has a multipicative effect on the hazard function. The model can be written as follows: h(t)=h 0 ( t )e x 1 p 1 + X 2 p 2 + + X p p p where h0(t) is the baseline hazard when all the X variables are equal to zero, and B's are the regression coefficients for the variable, and the log of Bj gives us the odds ratio associated 50 with each variable. The B's are estimated by a maximum likelihood method that does not depend on the shape of h(t) or h0(t), and the estimates measure the effect of each factor on the hazard function. Suppose there are K ordered and distinct onset times (ti.... tk). Let Rj = R(tj) represent the risk set at time tj (the number of individuals alive and uncensored at time ti). The partial likelihood for B is calculated with the following formula: L(B) = n k i = 1 e X ( i ) p 2-L£R(i) e where X(i) is the regression vector associated with the individual observed to have onset at time t|, and X(L) is the regression vector associated with all other individuals remaining at risk at time tj. Once the estimated values of the B coefficients are calculated, the likelihood ratio statistic can be used to test the null hypothesis that the B coefficients for all the variables of interest are simultaneously equal to zero. This test can be expressed as follows: LRS= 2 Log L(B) - 2 Log L fj) The likelihood ratio statistic can be solved by solving for B=0. The resulting value can be compared with the %2 distribution with n degrees of freedom (n = the total number of variables included in the model). By subtracting the deviance of a reduced model (model containing a reduced number of variables) from the deviance of the saturated model, we produce a measure of whether the reduced model can fit the data just as well as the saturated model. The resulting value is comparable to the x 2 distribution with n degrees of 51 freedom (n = the total number of variables in the saturated model - the total number of variables in the reduced model). All statistics concerning the Cox proportional hazard model were done with the EGRET statistical package (1993). Odds ratios were calculated separately for various subgroups of first degree relatives of AD cases and controls to determine if certain trends were specific to one group of first degree relatives, or if trends were similar in both groups. Odds ratios were also calculated for various subgroups of first degree relatives of AD cases and controls together, to determine if the risk to first degree relatives of AD cases is significantly greater than the risk to first degree relatives of controls. 4.5 Correction for the Gender Ratio in the General Elderly Population The risk to male first degree relatives of cases was adjusted for the Gender ratio found in CSHA (CSHA 1994 (a)) to determine if the significantly increased risk to female first degree relatives of cases is artifactual, reflecting the gender ratio observed in the general population, or if the risk to female first degree relatives of cases is actually greater than the risk to male first degree relatives of cases. It was only possible to do this correction for the part of our sample aged 65 and above, as CSHA gender ratio data were only available for individuals aged 65 and above. Two sex ratio corrections were used in the present study. These were 1.2:1 (CSHA ratio for the BC population), and 1.5:1 (CSHA ratio for the Canadian population). 52 5. Results 5.1 Samples Used for Analysis 5.1.1 C a s e s The diagnostic profile for the 1311 individuals assessed at the AD Clinic from 1985 to August 1995 is given in Table 5.1. Complete informative family history information was available for 463 index cases from families not believed to show an autosomal dominant pattern of inheritance. A total of 26 cases (5%) were excluded because an autosomal dominant pattern of inheritance was quite likely (see appendix B). Of these 463 index cases, 404 (87%) were diagnosed as probable AD, and 59 (13%) were diagnosed as autopsy confirmed AD at the time of this study. One hundred fifty-seven of these 463 index cases (34%) had early onset (age of onset <65 years) AD; 306 (66%) had late onset (age of onset > 65 years) AD. Data were available for a total of 866 and 1653 first degree relatives of early and late onset cases respectively. Table 5.2 shows the number and percentage of affected first degree relatives of the index cases by age of onset of the index case, gender of the relative, and relation between the index case and the first degree relative. The data on affected relatives in Table 5.2 are not age corrected. Age-specific risks for all sub-groups of first degree relatives were calculated up to age 88 years, which was the latest age for which the risk of AD could be calculated for sibs of early onset cases. 5.1.2 Controls Complete informative family histories were available for 796 consecutive unrelated controls ascertained through the CSHA and 4324 of their first degree relatives. Table 5.3 shows the number and percentage of affected first degree relatives of controls, by gender of the relative and relationship to index case. The data on affected relatives in Table 5.3 are not age corrected. Table 5.4 shows mean age and mean age at onset for cases and controls (early onset cases, late onset cases, and controls). 53 Table 5.1: Diagnosis for 1311 Alzheimer Clinic Patients Seen at the Alzheimer Clinic from 1985 to 1995, After Evaluation Clinic Diagnosis Number % of Total Demented, AD Unlikely 75 5.72 Demented, Possible AD 328 25.02 Demented, Probable AD 522 39.82 Autopsy Confirmed AD 79 6.02 Not Demented 290 22.12 Autopsy performed, not AD 17 1.30 Total 1311 100.00 Table 5.2: Number and Percent of Affected First Degree Relatives of Early and Late Onset Cases by Gender and Relation to Index Cases (not age corrected) Total No. No. Affected % Affected Early Onset Cases (N=157) Female relatives 441 13 2.95 Male relatives 425 5 1.18 Parents 318 17 5.35 Sibs 548 1 0.18 Late Onset Cases (N=306) Female relatives 816 27 3.31 Male relatives 837 12 1.43 Parents 589 20 3.40 Sibs 1064 19 1.79 Total Early Onset Cases 866 18 2.08 Late Onset Cases 1653 39 2.36 All Relatives 2519 57 2.26 54 Table 5.3: Number and Percent of Affected First Degree Relatives of Controls, by gender and relationship to controls (not age corrected) Relative Total No. No. Affected % Affected Females 2146 44 2.05 Males 2178 14 0.64 Parents 1579 36 2.28 Sibs 2745 22 0.80 All Relatives 4324 58 1.34 Table 5.4 : Early Onset Cases, Late Onset Cases, and Controls, by Gender, Mean Age, and Mean Age of Onset of Dementia. Early Onset Cases Late Onset Cases Controls 3 Number of Females 100 199 462 Number of Males 57 107 334 Mean Age 6 4 . 7 ± 6 . 3 7 6 . 2 ± 4 . 6 7 5 . 0 ± 6 . 3 Mean Age of Onset of Dementia 5 7 . 4 ± 5 . 3 7 1 . 6 ± 4 . 7 N/A a CSHA controls by definition were aged 65 and above 55 5.2 Risk Estimates for First Degree Relatives of Early Onset Cases , Late Onset C a s e s and Controls Kaplan-Meier age-specific risks for all first degree relatives, including parents and sibs, of early onset cases, late onset cases and controls are presented in Table 5.5. The cumulative lifetime risk up to age 88 for first degree relatives of early onset cases is 7.9+ 2.1%, compared to 8.0 ± 1.5% for first degree relatives of late onset cases and 4.1± 0.6% for first degree relatives of controls. The lifetime risk curves for these three groups of first degree relatives are plotted in Figure 5.1. The cumulative lifetime risk estimate to first degree relatives of early onset cases and late onset AD cases are non-significantly different from each other when compared with a difference of proportions test (Z=0.04, p=0.48), as are the lifetime risk curves when compared with a Log Rank test (Log Rank %2 = 0.41, p>0.10). The cumulative lifetime risk estimate for first degree relatives of late onset cases is significantly different from the cumulative lifetime risk estimate for first degree relatives of controls (Z=2.41, p<0.01), as are their respective lifetime risk curves (Log Rank x2 =9.38, p<0.005). The curve for first degree relatives of late onset cases increases at a greater rate compared to the curve for first degree relatives of controls, beginning at age 63 years. 56 Table 5.5 : Age-Specif ic Risks for First Degree Relatives of Early Onset C a s e s , Late Onset Cases , and Controls Relative Risk (%) to First Degree Risk (%) to First Degree Risk (%) to First Degree Age Relatives of Early Onset Relatives of Late Onset Relatives of Cont Cases + 95% C.I. Cases ± 95% C.I. + 95% C.I. 49 0.0 0.0 0.0 50 0.1 (0.0-1.0) 0.1 (0.0-0.5) 0.0 51 0.1 (0.0-1.0) 0.1 (0.0-0.5) 0.0 52 0.1 (0.0-1.0) 0.1 (0.0-0.5) 0.0 53 0.1 (0.0-1.0) 0.1 (0.0-0.5) 0.0 54 0.1 (0.0-1.0) 0.1 (0.0-0.5) 0.0 55 0.1 (0.0-1.0) 0.1 (0.0-0.5) 0.1 (0.0-0.2) 56 0.1 (0.0-1.0) 0.1 (0.0-0.5) 0.1 (0.0-0.2) 57 0.1 (0.0-1.0) 0.1 (0.0-0.5) 0.1 (0.0-0.2) 58 0.1 (0.0-1.0) 0.1 (0.0-0.5) 0.1 (0.0-0.2) 59 0.1 (0.0-1.0) 0.1 (0.0-0.6) 0.1 (0.0-0.2) 60 0.1 (0.0-1.0) 0.1 (0.0-0.6) 0.1 (0.0-0.3) 61 0.3 (0.1-1.3) 0.1 (0.0-0.6) 0.1 (0.0-0.3) 62 0.3 (0.1-1.3) 0.1 (0.0-0.6) 0.1 (0.0-0.3) 63 0.5 (0.2-1.6) 0.2 (0.1-0.7) 0.1 (0.0-0.3) 64 0.5 (0.2-1.6) 0.3 (0.1-0.8) 0.1 (0.1-0.3) 65 0.5 (0.2-1.6) 0.3 (0.1-0.8) 0.2 (0.1-0.4) 66 0.5 (0.2-1.6) 0.4 (0.2-0.9) 0.2 (0.1-0.4) 67 0.8 (0.3-2.0) 0.4 (0.2-0.9) 0.2 (0.1-0.5) 68 1.0 (0.4-2.5) 0.4 (0.2-0.9) 0.3 (0.1-0.5) 69 1.0 (0.4-2.5) 0.6 (0.3-1.2) 0.4 (0.2-0.7) 70 1.3 (0.6-3.0) 0.7 (0.3-1.4) 0.4 (0.2-0.8) 71 1.3 (0.6-3.0) 1.0 (0.6-1.9) 0.5 (0.3-0.8) 72 1.7 (0.8-3.6) 1.1 (0.6-2.0) 0.6 (0.4-1.0) 73 1.7 (0.8-3.6) 1.3 (0.7-2.2) 0.8 (0.5-1.3) 74 1.7 (0.8-3.6) 1.7 (1.0-2.7) 0.9 (0.6-1.4) 75 3.3 (1.8-6.2) 1.8 (1.1-2.9) 1.1 (0.7-1.6) 76 3.3 (1.8-6.2) 2.1 (1.3-3.4) 1.3 (0.9-2.0) 77 3.3 (1.8-6.2) 2.3 (1.5-3.6) 1.6 (1.1-2.2) 78 4.4 (2.5-7.8) 3.1 (2.0-4.7) 1.6 (1.1-2.3) 79 5.0 (2.9-8.7) 3.8 (2.5-5.6) 2.0 (1.4-2.7) 80 5.0 (2.9-8.7) 4.5 (3.1-6.5) 2.2 (1.6-3.1) 81 5.8 (3.4-9.8) 4.5 (3.1-6.5) 2.4 (1.8-3.3) 82 6.6 (3.9-11.1) 4.8 (3.3-7.0) 3.0 (2.2-4.0) 83 6.6 (3.9-11.1) 5.9 (4.1-8.5) 3.1 (2.3-4.2) 84 6.6 (3.9-11.1) 6.7 (4.7-9.6) 3.2 (2.4-4.4) 85 7.9 (4.6-13.3) 7.2 (5.0-10.3) 3.2 (2.4-4.4) 86 7.9 (4.6-13.3) 7.2 (5.0-10.3) 3.6 (2.7-4.9) 87 7.9 (4.6-13.3) 8.0 (5.5-11.6) 4.1 (3.0-5.6) 88 7.9 (4.6-13.3) 8.0 (5.5-11.6) 4.1 (3.0-5.6) 57 Figure 5.1 Age-Specific Risk (%) to First Degree Relatives of Early Onset Cases, Late Onset Cases, and Controls 14 12 + 10 + ^ f r L n m i o m i o c o c o c o i o t o h - r ^ h - ^ . ^ . c o o o c o c o AGE (years) Early Onset- Late Onset Controls • 95% confidence intervals for first degree relatives of early onset cases • 95% confidence intervals for first degree relatives of late onset cases — 95% confidence intervals for first degree relatives of controls 58 5.3 Risk Estimates for First Degree Relatives by Gender of the Relatives 5.3.1 Risk Estimates for Female and Male First Degree Relatives of Early Onset C a s e s Kaplan-Meier age-specific risks for female and male first degree relatives of early onset cases are presented in Table 5.6. The cumulative lifetime risk to age 88 years for female first degree relatives (10.8+ 3.2%) and male first degree relatives (3.5± 2.0%) of early onset cases differ significantly from each other (Z=1.93, p= 0.027). The risk curves (see Figure 5.2) suggest that female first degree relatives have a greater rate of increase in risk beginning at age 74 years, compared to male first degree relatives, but these do not differ significantly (Log Rank %2 = 2.50, p>0.10) 5.3.2 Risk Estimates for Female and Male First Degree Relatives of Late Onset C a s e s Kaplan-Meier age-specific risks for female and male first degree relatives of late onset cases are presented in Table 5.7. The cumulative lifetime risk to age 88 years for female (9.8± 2.2%) and male (5.3+ 1.8%) first degree relatives of late onset cases are not significantly different (Z=1.58, p=0.06). The lifetime risk curves of these two groups, plotted in Figure 5.3 show that both groups have similar increases in risk earlier in life, however female first degree relatives experience a greater increase in risk beginning at age 63 years, and continuing to age 88 years. These two lifetime risk curves do not differ significantly from each other (Log Rank %2= 3.57, p<0.10). 59 Table 5.6 : Age-Specif ic Risks for Female and Male First Degree Relatives of Early Onset Cases Relative Age Risk (%) to Female First Risk (%) to Male First Degree Relatives of Early Degree Relatives of Early Onset Cases + 95% C.I. Onset Cases + 95% C.I. 49 0.0 0.0 50 0.3 (0.0-1.8) 0.0 51 0.3 (0.0-1.8) 0.0 52 0.3 (0.0-1.8) 0.0 53 0.3 (0.0-1.8) 0.0 54 0.3 (0.0-1.8) 0.0 55 0.3 (0.0-1.8) 0.0 56 0.3 (0.0-1.8) 0.0 57 0.3 (0.0-1.8) 0.0 58 0.3 (0.0-1.8) 0.0 59 0.3 (0.0-1.8) 0.0 60 0.3 (0.0-1.8) 0.0 61 0.3 (0.0-1.8) 0.4 (0.1-2.7) 62 0.3 (0.0-1.8) 0.4 (0.1-2.7) 63 0.6 (0.2-2.5) 0.4 (0.1-2.7) 64 0.6 (0.2-2.5) 0.4 (0.1-2.7) 65 0.6 (0.2-2.5) 0.4 (0.1-2.7) 66 0.6 (0.2-2.5) 0.4 (0.1-2.7) 67 0.6 (0.2-2.5) 0.9 (0.2-3.8) 68 1.1 (0.3-3.4) 0.9 (0.2-3.8) 69 1.1 (0.3-3.4) 0.9 (0.2-3.8) 70 1.1 (0.3-3.4) 1.6 (0.5-5.2) 71 1.1 (0.3-3.4) 1.6 (0.5-5.2) 72 1.7 (0.6-4.6) 1.6 (0.5-5.2) 73 1.7 (0.6-4.6) 1.6 (0.5-5.2) 74 1.7 (0.6-4.6) 1.6 (0.5-5.2) 75 4.6 (2.2-9.2) 1.6 (0.5-5.2) 76 4.6 (2.2-9.2) 1.6 (0.5-5.2) 77 4.6 (2.2-9.2) 1.6 (0.5-5.2) 78 6.4 (3.4-11.9) 1.6 (0.5-5.2) 79 7.4 (4.0-13.4) 1.6 (0.5-5.2) 80 7.4 (4.0-13.4) 1.6 (0.5-5.2) 81 7.4 (4.0-13.4) 3.5 (1.1-11.1) 82 8.7 (4.8-15.5) 3.5 (1.1-11.1) 83 8.7 (4.8-15.5) 3.5 (1.1-11.1) 84 8.7 (4.8-15.5) 3.5 (1.1-11.1) 85 10.8 (5.9-19.3) 3.5 (1.1-11.1) 86 10.8 (5.9-19.3) 3.5 (1.1-11.1) 87 10.8 (5.9-19.3) 3.5 (1.1-11.1) 88 10.8 (5.9-19.3) 3.5 (1.1-11.1) 60 Figure 5.2: Age-Specific Risk (%) to Female and Male First Degree Relatives of Early Onset Cases 20 j -18 -16 --14 + O C M ^ - ( 0 ' 0 0 O C M ^ - < D O O C \ l ^ - C 0 0 0 to us <o <o co ^ N N oo oo co ' co oo AGE (years) Females Males • 95% confidence intervals for female first degree relatives of early onset cases — 95% confidence intervals for male first degree relatives of early onset cases 61 Table 5.7 : Age-Specif ic Risks for Female and Male First Degree Relatives of Late Onset C a s e s Relative Age Risk (%) to Female First Risk (%) to Male First Degree Relatives of Late Degree Relatives of Late Onset Cases ± 95% C.I. Onset Cases ± 95% C.I. 49 0.0 0.0 50 0.1 (0.0-1.0) 0.0 51 0.1 (0.0-1.0) 0.0 52 0.1 (0.0-1.0) 0.0 53 0.1 (0.0-1.0) 0.0 54 0.1 (0.0-1.0) 0.0 55 0.1 (0.0-1.0) 0.0 56 0.1 (0.0-1.0) 0.0 57 0.1 (0.0-1.0) 0.0 58 0.1 (0.0-1.0) 0.0 59 0.1 (0.0-1.0) 0.2 (0.0-1.1) 60 0.1 (0.0-1.0) 0.2 (0.0-1.1) 61 0.1 (0.0-1.0) 0.2 (0.0-1.1) 62 0.1 (0.0-1.0) 0.2 (0.0-1.1) 63 0.3 (0.1-1.2) 0.2 (0.0-1.1) 64 0.4 (0.1-1.4) 0.2 (0.0-1.1) 65 0.4 (0.1-1.4) 0.2 (0.0-1.1) 66 0.6 (0.2-1.6) 0.2 (0.0-1.1) 67 0.6 (0.2-1.6) 0.2 (0.0-1.1) 68 0.6 (0.2-1.6) 0.2 (0.0-1.1) 69 1.0 (0.4-2.2) 0.2 (0.0-1.1) 70 1.2 (0.6-2.5) 0.2 (0.0-1.1) 71 1.6 (0.8-3.1) 0.4 (0.1-1.6) 72 1.8 (1.0-3.3) 0.4 (0.1-1.6) 73 2.0 (1.1-3.7) 0.4 (0.1-1.6) 74 2.3 (1.3-4.0) 1.0 (0.4-2.6) 75 2.3 (1.3-4.0) 1.3 (0.5-3.1) 76 2.9 (1.7-4.9) 1.3 (0.5-3.1) 77 3.2 (1.9-5.3) 1.3 (0.5-3.1) 78 3.5 (2.1-5.8) 2.7 (1.3-5.4) 79 4.3 (2.7-6.9) 3.2 (1.6-6.3) 80 5.5 (3.5-8.5) 3.2 (1.6-6.3) 81 5.5 (3.5-8.5) 3.2 (1.6-6.3) 82 6.0 (3.9-9.2) 3.2 (1.6-6.3) 83 7.2 (4.7-10.9) 4.2 (2.1-8.4) 84 7.8 (5.1-11.8) 5.3 (2.7-10.4) 85 8.6 (5.6-12.9) 5.3 (2.7-10.4) 86 8.6 (5.6-12.9) 5.3 (2.7-10.4) 87 9.8 (6.3-14.9) 5.3 (2.7-10.4) 88 9.8 (6.3-14.9) 5.3 (2.7-10.4) 62 Figure 5.3: Age-Specific Risk (%) to Female and Male First Degree Relatives of Late Onset Cases 2 + 8 + 6 + 4 + 2 + 0 T—'I I1 I M I I I I1 0 5 i - c o i r ) t ^ O ) i - c o m r ^ O ) T - c o i o N c n ^ c o i n h . ^ - i o i o m m m < o < o c o < o . ( O N h > . N N h « c o c o c o o o AGE (years) Females Males • 95% confidence intervals for female first degree relatives of late onset cases — 95% confidence intervals for male first degree relatives of late onset cases 63 5.3.3 Risk Estimates for Female and Male First Degree Relatives of Controls Kaplan-Meier age-specific risks for female and male first degree relatives of controls are presented in Table 5.8. The cumulative lifetime risk to age 88 years for female first degree relatives (5.1 ± 0.9%) and male first degree relatives (2.6± 0.8%) of controls are significantly different (Z=2.08, p=0.019). The lifetime risk curves plotted in Figure 5.4 show that the risk to female first degree relatives has a tendency to increase at a significantly greater rate than the one for male first degree relatives beginning at age 67 (Log Rank X2=6.95, p<0.01). 5.4 Risk Estimates for First Degree Relatives by Relationship to the Index Case 5.4.1 Risk Estimates for Parents and Sibs of Early Onset Cases Kaplan-Meier age-specific risks for sibs and parents of early onset cases are presented in Table 5.9. The cumulative lifetime risk to age 88 years for parents (9.9 ± 2.5%) and sibs (1.1± 1.0%) of early onset cases are significantly different from each other (Z= 3.27 p<0.001). The risk curves of these two groups as shown in Figure 5.5 are also significantly different from each other (Log Rank %2= 6.21, p<0.025). The risk to parents increases at an earlier age, and rises continually over time, while the one affected sib has onset at age 72, resulting in one peak and subsequent plateau. 5.4.2 Risk Estimates for Parents and Sibs of Late Onset Cases Kaplan-Meier age-specific risks for sibs and parents of late onset cases are presented in Table 5.10. The cumulative lifetime risk to age 88 years for parents (7.7± 1.8%) and sibs (9.0± 3.2%) of late onset cases do not differ significantly (Z=0.35, p=0.36). The two lifetime risk curves (Figure 5.6) are also not significantly different (Log Rank x2 = 0.06, p>0.10). 64 Table 5.8 : Age-Specif ic Risks for Female and Male First Degree Relatives of Controls Relative Age Risk (%) to Female First Risk (%) to Male First Degree Relatives of Degree Relatives of Controls ± 95% C.I. Controls + 95% C.I. 54 0.0 0.0 55 0.1 (0.0-0.4) 0.1 (0.0-0.4) 56 0.1 (0.0-0.4) 0.1 (0.0-0.4) 57 0.1 (0.0-0.4) 0.1 (0.0-0.4) 58 0.1 (0.0-0.4) 0.1 (0.0-0.4) 59 0.1 (0.0-0.4) 0.1 (0.0-0.4) 60 0.1 (0.0-0.4) 0.1 (0.0-0.4) 61 0.1 (0.0-0.4) 0.1 (0.0-0.4) 62 0.1 (0.0-0.4) 0.1 (0.0-0.4) 63 0.1 (0.0-0.4) 0.1 (0.0-0.5) 64 0.1 (0.0-0.4) 0.1 (0.0-0.5) 65 0.2 (0.1-0.5) 0.1 (0.0-0.5) 66 0.2 (0.1-0.6) 0.1 (0.0-0.5) 67 0.3 (0.1-0.7) 0.1 (0.0-0.5) 68 0.4 (0.2-0.8) 0.1 (0.0-0.5) 69 0.5 (0.3-1.0) 0.2 (0.1-0.7) 70 0.6 (0.3-1.2) 0.2 (0.1-0.7) 71 0.7 (0.4-1.3) 0.2 (0.1-0.7) 72 0.8 (0.5-1.5) 0.3 (0.1-0.9) 73 1.2 (0.7-2.0) 0.3 (0.1-0.9) 74 1.4 (0.9-2.2) 0.3 (0.1-0.9) 75 1.6 (1.0-2.5) 0.5 (0.2-1.1) 76 1.9 (1.3-2.9) 0.6 (0.3-1.4) 77 2.3 (1.6-3.4) 0.6 (0.3-1.4) 78 2.4 (1.7-3.5) 0.6 (0.3-1.4) 79 2.6 (1.8-3.7) 1.2 (0.6-2.4) 80 2.9 (2.0-4.1) 1.4 (0.7-2.8) 81 3.2 (2.2-4.5) 1.4 (0.7-2.8) 82 3.7 (2.6-5.2) 2.1 (1.1-3.8) 83 3.9 (2.8-5.5) 2.1 (1.1-3.8) 84 4.1 (2.9-5.7) 2.1 (1.1-3.8) 85 4.1 (2.9-5.7) 2.1 (1.1-3.8) 86 4.4 (3.1-6.2) 2.6 (1.4-5.0) 87 5.1 (3.6-7.1) 2.6 (1.4-5.0) 88 5.1 (3.6-7.1) 2.6 (1.4-5.0) 65 Figure 5.4: Age-Specific Risk (%) to Female and Male First Degree Relatives of Controls 8 -r-7 -6 --" 5 t - C O O O O C \ l " 3 - C D O O O C \ | - 5 r < O C O O C \ | - < l - C O C O L O l O l O ( D < O ( O < D ( O h - l ^ r - r ' - f - - 0 0 C 0 0 0 0 0 C0 AGE (years) Females Males • 95% confidence intervals for female first degree relatives of controls — 95% confidence intervals for male first degree relatives of controls 66 Table 5.9 : Age-Specif ic Risks for Sibs and Parents of Early Onset Cases Relative Age Risk (%) to Sibs of Early Risk (%) to Parents of Onset Cases ± 95% C.I. Early Onset Cases ± 95% CA. 49 0.0 0.0 50 0.0 0.3 (0.0-2.4) 51 0.0 0.3 (0.0-2.4) 52 0.0 0.3 (0.0-2.4) 53 0.0 0.3 (0.0-2.4) 54 0.0 0.3 (0.0-2.4) 55 0.0 0.3 (0.0-2.4) 56 0.0 0.3 (0.0-2.4) 57 0.0 0.3 (0.0-2.4) 58 0.0 0.3 (0.0-2.4) 59 0.0 0.3 (0.0-2.4) 60 0.0 0.3 (0.0-2.4) 61 0.0 0.7 (0.2-2.9) 62 0.0 0.7 (0.2-2.9) 63 0.0 1.1 (0.4-3.4) 64 0.0 1.1 (0.4-3.4) 65 0.0 1.1 (0.4-3.4) 66 0.0 1.1 (0.4-3.4) 67 0.0 1.6 (0.6-4.1) 68 0.0 2.0 (0.8-4.8) 69 0.0 2.0 (0.8-4.8) 70 0.0 2.5 (1.1-5.5) 71 0.0 2.5 (1.1-5.5) 72 1.1 (0.2-7.4) 2.5 (1.1-5.5) 73 1.1 (0.2-7.4) 2.5 (1.1-5.5) 74 1.1 (0.2-7.4) 2.5 (1.1-5.5) 75 1.1 (0.2-7.4) 4.8 (2.6-8.8) 76 1.1 (0.2-7.4) 4.8 (2.6-8.8) 77 1.1 (0.2-7.4) 4.8 (2.6-8.8) 78 1.1 (0.2-7.4) 6.1 (3.5-10.7) 79 1.1 (0.2-7.4) 6.9 (4.0-11.7) 80 1.1 (0.2-7.4) 6.9 (4.0-11.7) 81 1.1 (0.2-7.4) 7.7 (4.6-13.0) 82 1.1 (0.2-7.4) 8.6 (5.2-14.6) 83 1.1 (0.2-7.4) 8.6 (5.2-14.6) 84 1.1 (0.2-7.4) 8.6 (5.2-14.6) 85 1.1 (0.2-7.4) 9.9 (6.0-16.3) 86 1.1 (0.2-7.4) 9.9 (6.0-16.3) 87 1.1 (0.2-7.4) 9.9 (6.0-16.3) 88 1.1 (0.2-7.4) 9.9 (6.0-16.3) 67 Figure 5.5: Age-Specific Risk (%) to Sibs and Parents of Early Onset Cases 16 14 1 2 1 0 + 52 8 i i i i i i i i I i i i i i i i r i i i . -t-+-r»- a> T-C0 CO N AGE (years) co i n O) T- co i n N oo co oo co -Sibs - Parents • 95% confidence intervals for sibs of early onset cases — 95% confidence intervals for parents of early onset cases 68 Table 5.10 : Age-Specif ic Risks for Sibs and Parents of Late Onset Cases Relative Age Risk (%) to Sibs of Late Risk (%) to Parents of Onset Cases ± 95% C.I. Late Onset C a s e s ± 95% C.I. 49 0.0 0.0 50 0.0 0.2 (0.0-1.4) 51 0.0 0.2 (0.0-1.4) 52 0.0 0.2 (0.0-1.4) 53 0.0 0.2 (0.0-1.4) 54 0.0 0.2 (0.0-1.4) 55 0.0 0.2 (0.0-1.4) 56 0.0 0.0 (0.0-0.0) 57 0.0 0.2 (0.0-1.4) 58 0.0 0.2 (0.0-1.4) 59 0.1 (0.0-0.8) 0.2 (0.0-1.4) 60 0.1 (0.0-0.8) 0.2 (0.0-1.4) 61 0.1 (0.0-0.8) 0.2 (0.0-1.4) 62 0.1 (0.0-0.8) 0.2 (0.0-1.4) 63 0.1 (0.0-0.8) 0.4 (0.1-1.6) 64 0.2 (0.1-1.0) 0.4 (0.1-1.6) 65 0.2 (0.1-1.0) 0.4 (0.1-1.6) 66 0.2 (0.1-1.0) 0.6 (0.2-2.0) 67 0.2 (0.1-1.0) 0.6 (0.2-2.0) 68 0.2 (0.1-1.0) 0.6 (0.2-2.0) 69 0.6 (0.2-1.5) 0.6 (0.2-2.0) 70 0.6 (0.2-1.5) 0.9 (0.3-2.4) 71 0.8 (0.3-1.8) 1.4 (0.6-3.2) 72 1.0 (0.4-2.1) 1.4 (0.6-3.2) 73 1.2 (0.6-2.5) 1.4 (0.6-3.2) 74 1.4 (0.7-2.8) 2.0 (1.0-4.1) 75 1.4 (0.7-2.8) 2.3 (1.2-4.5) 76 1.7 (0.9-3.4) 2.7 (1.4-5.0) 77 2.1 (1.1-4.0) 2.7 (1.4-5.0) 78 3.7 (2.1-6.4) 2.7 (1.4-5.0) 79 5.1 (3.1-8.4) 2.7 (1.4-5.0) 80 5.1 (3.1-8.4) 4.0 (2.3-7.0) 81 5.1 (3.1-8.4) 4.0 (2.3-7.0) 82 5.1 (3.1-8.4) 4.5 (2.7-7.7) 83 6.1 (3.6-10.3) 5.7 (3.4-9.4) 84 6.1 (3.6-10.3) 6.9 (4.3-11.1) 85 6.1 (3.6-10.3) 7.7 (4.8-12.2) 86 6.1 (3.6-10.3) 7.7 (4.8-12.2) 87 9.0 (4.4-18.2) 7.7 (4.8-12.2) 88 9.0 (4.4-18.2) 7.7 (4.8-12.2) 69 Figure 5.6: Age-Specific Risk (%) to Sibs and Parents of Late Onset Cases 18 --16 --14 12 + c n i - c o u i r ^ c n ^ c o i n r ^ o i T - c o i r j r ^ c n - i - c o i o r ^ ^ • l O i o i o i o i o c o t o t o c o c o i ^ t ^ ^ N t ^ c o o o o o e o AGE (years) Sibs Parents • 95% confidence intervals for sibs of late onset cases — 95% confidence intervals for parents of late onset cases 70 5.4.3 Risk Estimates for Parents and Sibs of Controls Kaplan-Meier age-specific risks for sibs and parents of controls are presented in Table 5.11. The cumulative lifetime risk to age 88 years for parents (4.1+ 0.8%) and sibs (4.8± 1.6%) of controls are non-significantly different (Z=0.22, p=0.41). The lifetime risk curves of parents and sibs of controls are plotted in Figure 5.7. The curves show that the risk to both groups remain relatively similar up to the age of 88 years, and therefore the risk curves to parents and sibs of controls are not significantly different (Log Rank %2= 0.73, p>0.10). 5.5 Risk Estimates for First Degree Relatives by Gender of the Proband 5.5.1 Risk Estimates for First Degree Relatives of Female and Male Early Onset C a s e s Kaplan-Meier age-specific risks for first degree relatives of female and male early onset cases are presented in Table 5.12. The cumulative lifetime risk to age 88 years for first degree relatives of female early onset cases (6.6± 2.2%) is not significantly different from the cumulative lifetime risk to first degree relatives of male early onset cases (10.6± 4.7%) (Z = 0.77, p=0.22). The lifetime risk curves for first degree relatives of female and male early onset cases are plotted in Figure 5.8. The two curves do not differ significantly (Log Rank %2= 0.72, p>0.10). 71 Table 5.11 : Age-Specific Risks for Sibs and Parents of Controls Relative Age Risk (%) to Sibs of Risk (%) to Parents of Controls + 95% C.I. Controls + 95% C.I. 54 0.0 0.0 55 0.0 0.1 (0.0-0.5) 56 0.0 0.1 (0.0-0.5) 57 0.0 0.1 (0.0-0.5) 58 0.0 0.1 (0.0-0.5) 59 0.0 0.1 (0.0-0.5) 60 0.1 (0.1-0.4) 0.1 (0.0-0.5) 61 0.1 (0.1-0.4) 0.1 (0.0-0.5) 62 0.1 (0.1-0.4) 0.1 (0.0-0.5) 63 0.1 (0.1-0.5) 0.1 (0.0-0.5) 64 0.1 (0.1-0.5) 0.2 (0.0-0.6) 65 0.1 (0.1-0.5) 0.2 (0.0-0.6) 66 0.2 (0.1-0.6) 0.2 (0.0-0.6) 67 0.2 (0.1-0.6) 0.2 (0.1-0.8) 68 0.3 (0.1-0.7) 0.2 (0.1-0.8) 69 0.3 (0.2-0.8) 0.4 (0.2-1.2) 70 0.3 (0.2-0.8) 0.5 (0.2-1.2) 71 0.3 (0.2-0.8) 0.6 (0.3-1.3) 72 0.5 (0.3-1.1) 0.7 (0.4-1.4) 73 0.5 (0.3-1.1) 1.1 (0.6-2.0) 74 0.6 (0.3-1.2) 1.3 (0.7-2.2) 75 0.9 (0.5-1.6) 1.4 (0.8-2.3) 76 1.1 (0.6-2.0) 1.6 (1.0-2.6) 77 1.3 (0.7-2.2) 1.9 (1.2-2.9) 78 1.3 (0.7-2.2) 2.0 (1.3-3.1) 79 1.7 (1.0-2.9) 2.3 (1.5-3.5) 80 1.7 (1.0-2.9) 2.7 (1.8-4.0) 81 2.0 (1.1-3.3) 2.9 (1.9-4.2) 82 2.9 (1.7-4.9) 3.2 (2.2-4.7) 83 2.9 (1.7-4.9) 3.4 (2.3-4.9) 84 2.9 (1.7-4.9) 3.6 (2.5-5.2) 85 2.9 (1.7-4.9) 3.6 (2.5-5.2) 86 3.8 (2.1-6.8) 3.8 (2.6-5.5) 87 4.8 (2.6-8.9) 4.1 (2.8-5.9) 88 4.8 (2.6-8.9) 4.1 (2.8-5.9) 72 Figure 5.7: Age-Specific Risk (%) to Sibs and Parents of Controls r? 5 + co DC 4 AGE (years) -Sibs - Parents • 95% confidence intervals for sibs of controls — 95% confidence intervals for parents of controls 73 Table 5.12 : Age-Specif ic Risks for First Degree Relatives of Female and Male Early Onset Cases Relative Age Risk (%) to First Degree Risk (%) to First Degree Relatives of Female Early Relatives of Male Early Onset Cases + 95% C.I. Onset Cases + 95% C.I. 49 0.0 0.0 50 0.2 (0.0-1.5) 0.0 51 0.2 (0.0-1.5) 0.0 52 0.2 (0.0-1.5) 0.0 53 0.2 (0.0-1.5) 0.0 54 0.2 (0.0-1.5) 0.0 55 0.2 (0.0-1.5) 0.0 56 0.2 (0.0-1.5) 0.0 57 0.2 (0.0-1.5) 0.0 58 0.2 (0.0-1.5) 0.0 59 0.2 (0.0-1.5) 0.0 60 0.2 (0.0-1.5) 0.0 61 0.5 (0.1-1.9) 0.0 62 0.5 (0.1-1.9) 0.0 63 0.5 (0.1-1.9) 0.6 (0.1-4.1) 64 0.5 (0.1-1.9) 0.6 (0.1-4.1) 65 0.5 (0.1-1.9) 0.6 (0.1-4.1) 66 0.5 (0.1-1.9) 0.6 (0.1-4.1) 67 0.5 (0.1-1.9) 1.4 (0.3-5.4) 68 0.5 (0.1-1.9) 2.2 (0.7-6.4) 69 0.5 (0.1-1.9) 2.2 (0.7-6.4) 70 0.9 (0.3-3.0) 2.2 (0.7-6.4) 71 0.9 (0.3-3.0) 2.2 (0.7-6.4) 72 0.9 (0.3-3.0) 3.3 (1.2-8.6) 73 0.9 (0.3-3.0) 3.3 (1.2-8.6) 74 0.9 (0.3-3.0) 3.3 (1.2-8.6) 75 2.7 (1.2-6.3) 4.7 (1.9-11.3) 76 2.7 (1.2-6.3) 4.7 (1.9-11.3) 77 2.7 (1.2-6.3) 4.7 (1.9-11.3) 78 4.3 (2.1-8.8) 4.7 (1.9-11.3) 79 4.3 (2.1-8.8) 6.6 (2.8-10.5) 80 4.3 (2.1-8.8) 6.6 (2.8-10.5) 81 5.4 (2.7-10.7) 6.6 (2.8-10.5) 82 6.6 (3.4-12.8) 6.6 (2.8-10.5) 83 6.6 (3.4-12.8) 6.6 (2.8-10.5) 84 6.6 (3.4-12.8) 6.6 (2.8-10.5) 85 6.6 (3.4-12.8) 10.6 (4.3-24.9) 86 6.6 (3.4-12.8) 10.6 (4.3-24.9) 87 6.6 (3.4-12.8) 10.6 (4.3-24.9) 88 6.6 (3.4-12.8) 10.6 (4.3-24.9) 74 Figure 5.8: Age-Specific Risk (%) to First Degree Relatives of Female and Male Early Onset Cases 25 7 20 + 1 5 -10 --5 -M - m m m i r > m c 0 t o t o ( D ( D r - - h - r - - K r ^ - c o o o c o c o AGE (years) Female Cases Male Cases • 95% confidence intervals for first degree relatives of female early onset cases — 95% confidence intervals for first degree relatives of male early onset cases 75 5.5.2 Risk Estimates for First Degree Relatives of Female and Male Late Onset C a s e s Kaplan-Meier age-specific risks for first degree relatives of female and male late onset cases are presented in Table 5.13. The cumulative lifetime risk to age 88 years for first degree relatives of female late onset cases (6.4± 1.5%) is not significantly different from that for first degree relatives of male late onset cases (10.7± 3.0%) when compared with a difference of proportions test (Z=1.28, p=0.10). The risk to both groups of first degree relatives remain similar to age 78, when the risk to first degree relatives of male late onset cases begins to increase at a greater rate. Nevertheless, both lifetime risk curves (Figure 5.9) are not significantly different (Log Rank %2 = 0.93, p>0.10). 5.5.3 Risk Estimates for First Degree Relatives of Female and Male Controls Kaplan-Meier age-specific risks for first degree relatives of female and male controls are presented in Table 5.14. The cumulative lifetime risk to age 88 years for first degree relatives of female controls (3.7± 0.8%) is not significantly different from the cumulative lifetime risk to first degree relatives of male controls (4.6± 1.1%) (Z = 0.66, p=0.25). The two risk curves are plotted in Figure 5.10. Both risk curves follow a similar pattern of increasing risk with advancing age, and are not significantly different (Log Rank %2 = 0.31, p>0.10). 76 Table 5.13 : Age-Specif ic Risks for First Degree Relatives of Female and Male Late Onset Cases Relative Age Risk (%) to First Degree Risk (%) to First Degree Relatives of Female Late Relatives of Male Late Onset Cases ± 95% C.I. Onset Cases ± 95% C.I. 49 0.0 0.0 50 0.1 (0.0-0.8) 0.0 51 0.1 (0.0-0.8) 0.0 52 0.1 (0.0-0.8) 0.0 53 0.1 (0.0-0.8) 0.0 54 0.1 (0.0-0.8) 0.0 55 0.1 (0.0-0.8) 0.0 56 0.1 (0.0-0.8) 0.0 57 0.1 (0.0-0.8) 0.0 58 0.1 (0.0-0.8) 0.0 59 0.2 (0.1-0.9) 0.0 60 0.2 (0.1-0.9) 0.0 61 0.2 (0.1-0.9) 0.0 62 0.2 (0.1-0.9) 0.0 63 0.2 (0.1-0.9) 0.2 (0.0-1.6) 64 0.3 (0.1-1.1) 0.2 (0.0-1.6) 65 0.3 (0.1-1.1) 0.2 (0.0-1.6) 66 0.3 (0.1-1.1) 0.5 (0.1-1.9) 67 0.3 (0.1-1.1) 0.5 (0.1-1.9) 68 0.3 (0.1-1.1) 0.5 (0.1-1.9) 69 0.6 (0.3-1.6) 0.5 (0.1-1.9) 70 0.8 (0.4-1.8) 0.5 (0.1-1.9) 71 1.1 (0.6-2.3) 0.8 (0.3-2.5) 72 1.3 (0.7-2.6) 0.8 (0.3-2.5) 73 1.5 (0.8-2.8) 0.8 (0.3-2.5) 74 1.5 (0.8-2.8) 1.9 (0.9-4.3) 75 1.7 (0.9-3.1) 1.9 (0.9-4.3) 76 2.2 (1.3-3.9) 1.9 (0.9-4.3) 77 2.2 (1.3-3.9) 2.5 (1.1-5.2) 78 2.8 (1.7-4.8) 3.6 (1.8-7.0) 79 3.2 (1.9-5.3) 4.9 (2.6-8.9) 80 3.9 (2.4-6.4) 5.6 (3.1-9.9) 81 3.9 (2.4-6.4) 5.6 (3.1-9.9) 82 3.9 (2.4-6.4) 6.5 (3.7-11.4) 83 5.1 (3.1-8.2) 7.6 (4.3-13.1) 84 5.7 (3.5-9.1) 8.7 (5.1-14.9) 85 6.4 (4.0-10.3) 8.7 (5.1-14.9) 86 6.4 (4.0-10.3) 8.7 (5.1-14.9) 87 6.4 (4.0-10.3) 10.7 (6.1-18.6) 88 6.4 (4.0-10.3) 10.7 (6.1-18.6) 77 Figure 5.9: Age-Specific Risk (%) to First Degree Relatives of Female and Male Late Onset Cases 18 + 16 + 14 + 12 + 10 --to DC 8 --6 + 4 + 2 + 0 f I I I I I I I i I I I l ~ F T T T I I o ^ T - c o i o r ^ c J i - t - c o i n r ^ c n T - c o i n i ^ c n - r - c o i o i ^ -^ t m m i f ) i r ) i o c o c 0 < o c D ( D h - h - h - i - - r ^ . o o o o o o o o AGE (years) I Female Cases Male Cases I • 95% confidence intervals for first degree relatives of female late onset cases — 95% confidence intervals for first degree relatives of male late onset cases 78 Table 5.14 : Age-Specif ic Risks for First Degree Relatives of Female and Male Controls Relative Age Risk (%) to First Degree Risk (%) to First Degree Relatives of Female Relatives of Male Controls ± 95% C.I. Controls ± 95% C.I. 53 0.0 0.0 54 0.0 0.1 (0.0-0.5) 55 0.0 0.1 (0.0-0.5) 56 0.0 0.1 (0.0-0.5) 57 0.0 0.1 (0.0-0.5) 58 0.0 0.1 (0.0-0.5) 59 0.0 0.1 (0.0-0.5) 60 0.1 (0.0-0.4) 0.1 (0.0-0.5) 61 0.1 (0.0-0.4) 0.1 (0.0-0.5) 62 0.1 (0.0-0.4) 0.1 (0.0-0.5) 63 0.1 (0.0-0.4) 0.1 (0.0-0.6) 64 0.1 (0.0-0.4) 0.1 (0.0-0.6) 65 0.2 (0.0-0.5) 0.1 (0.0-0.6) 66 0.2 (0.0-0.5) 0.2 (0.1-0.7) 67 0.2 (0.0-0.5) 0.3 (0.1-0.9) 68 0.2 (0.0-0.5) 0.3 (0.1-0.9) 69 0.4 (0.2-0.9) 0.3 (0.1-0.9) 70 0.5 (0.2-1.0) 0.3 (0.1-0.9) 71 0.6 (0.3-1.1) 0.3 (0.1-0.9) 72 0.8 (0.4-1.4) 0.3 (0.1-0.9) 73 1.0 (0.6-1.6) 0.6 (0.3-1.3) 74 1.1 (0.6-1.7) 0.7 (0.3-1.5) 75 1.1 (0.7-1.9) 1.0 (0.5-2.0) 76 1.3 (0.8-2.0) 1.5 (0.8-2.7) 77 1.5 (0.9-2.3) 1.7 (1.0-2.9) 78 1.6 (1.0-2.5) 1.7 (1.0-2.9) 79 1.9 (1.2-2.9) 2.1 (1.2-3.5) 80 2.2 (1.4-3.3) 2.3 (1.4-3.8) 81 2.3 (1.6-3.5) 2.6 (1.6-4.2) 82 2.7 (1.8-4.1) 3.4 (2.1-5.4) 83 2.7 (1.8-4.1) 3.7 (2.3-5.8) 84 2.7 (1.8-4.1) 4.1 (2.6-6.3) 85 2.7 (1.8-4.1) 4.1 (2.6-6.3) 86 3.3 (2.2-5.0) 4.1 (2.6-6.3) 87 3.7 (2.4-5.6) 4.6 (2.9-7.3) 88 3.7 (2.4-5.6) 4.6 (2.9-7.3) 79 Figure 5.10: Age-Specific Risk (%) to First Degree Relatives of Female and Male Controls 8 y 7 6 -5 -T t ( 0 0 0 O C V I ' 4 - ( 0 C D O C M T t C 0 C 0 O ( M ' 4 - ( 0 0 0 1 Q I O I O ( 0 . ( O C O I O C O N ^ N N ^ C O O O C O C O C O AGE (years) Female controls Male Controls ( • 95% confidence intervals for first degree relatives of female controls — 95% confidence intervals for first degree relatives of male controls 80 5.6 Summary of Risk Estimates for Various Subgroups of First Degree Relatives Comparison groups Difference of Proportions p value Log Rank p value Age of onset of the proband First Degree Relatives of Early vs Late Onset Cases First Degree Relatives of Late Onset Cases vs Controls .48 <.01 >.10 <.005 Gender of the Relative Female vs Male First Degree Relatives of Early Onset Cases Female vs Male First Degree Relatives of Late Onset Cases Female vs Male First Degree Relatives of Controls .027 .06 .019 >.10 <.10 <.01 Relation Parents vs Sibs of Early Onset Cases Parents vs Sibs of Late Onset Cases Parents vs Sibs of Controls <.001 .36 .41 <.025 >.10 >.10 Gender of the Proband First Degree Relatives of Female vs Male Early Onset Cases First Degree Relatives of Female vs Male Late Onset Cases First Degree Relatives of Female vs Male Controls .22 .10 .25 >.10 >.10 >.10 81 5.7 Relative Risk Odds Ratios for First Degree Relatives of Early Onset C a s e s and Late Onset Cases by Gender, and Relationship to Index Case Odds ratios for various subgroups of first degree relatives of early onset cases and late onset cases, calculated with the Cox proportional hazards model, are presented in Tables 5.15 and 5.16. The first model tested (Table 5.15) includes the following variables of interest: (i) age of onset of the index case, (ii) gender of the index case, (iii) relationship to the index case, and (iv) gender of the relative. This model produced a significant odds ratio for gender of the relative. The odds ratio for female first degree relatives of cases was 1.981 (95% C.l.= 1.110- 3.536), in relation to male first degree relatives of all cases. The results of the reduced model are presented in Table 5.16. The gender of the first degree relative appeared to be a significant risk modifying factor, with female first degree relatives having an odds ratio of 1.959 in relation to male first degree relatives (95% C.I. = 1.098 - 3.495). The difference between these two models is not significant when compared with a likelihood ratio test (LRS = 4.055 with df=3, p>0.05) Table 5.15: Odds Ratios for Various Subgroups of First Degree Relatives of AD Cases Variable Odds Ratio 95% C.I. Age of Onset of Proband Late Onset 1.00 Early Onset 1.157 .6530-2.049 Gender of the Proband Female 1.00 Male 1.270 .8777-1.838 Relation to the Proband Sibs 1.00 Parents 1.466 .8155-2.563 Gender of the Relative Male 1.00 Female 1.981 1.110-3.536 Deviance = 743.901 Likelihood Ratio Statistic on 4 df = 9.689 p=0.046 8 2 Table 5.16: Odds Ratio for Female First Degree Relatives of AD Cases (Reduced Model) Variable Odds Ratio 95% C.I. Gender of the Relative Female 1.00 Male 1.959 1.098-3.495 Deviance = 747.956 Likelihood Ratio Statistic on 1 df = 5.634 p=0.018 5.8 Relative Risk Odd Ratios for First Degree Relatives of Controls by Gender, and Relationship to Index Case Odds ratios for various subgroups of first degree relatives of controls, calculated with the Cox proportional hazards model, are presented in Tables 5.17 and 5.18. The first model tested (Table 5.17) includes three variables of interest: (i) gender of the index case, (ii) relationship to the index case, and (iii) gender of the relative. The odds ratios for all variables other than sex of the relative did not differ significantly. The odds ratio for female first degree relatives was 2.222 (95% C.I. = 1.215-4.061), in relation to male first degree relatives of controls. This is in agreement with the results of section 5.6, in which the odds ratio for female first degree relatives of cases was 1.857 in relation to male first degree relatives. The results of the reduced model are presented in Table 5.18. The gender of the first degree relative proved to be a significant risk modifying factor, with female first degree relatives having an odds ratio of 2.207 in relation to male first degree relatives (95% C.I. = 1.207 - 4.034). The difference between these two models is non-significant when compared with a likelihood ratio test (LRS = 1.214 with df = 2, p>0.05) 83 Table 5.17: Odds Ratios for Various Subgroups of First Degree Relatives of Controls Variable Odds Ratio 95% C.I. Gender of the Proband Male 1.00 Female .8610 0.5116-1.449 Relationship to Proband Sibs 1.00 Parents 1.273 0.7348-2.206 Gender of the Relative Male 1.00 Female 2.222 1.215-4.061 Deviance = 816.574 Likelihood Ratio Statistic on 3 df = 8.655 p=0.036 Table 5.18: Odds Ratio for Female First Degree Relatives of Controls (Reduced Model) Variable Odds Ratio 95% C.I. Gender of the Relative Male 1.00 Female 2.207 1.207-4.034 Deviance = 817.681 Likelihood Ratio Statistic on 1 df = 7.441 p=0.006 84 5.9 Relative Risk Odds Ratios for First Degree Relatives of Early Onset Cases, Late Onset Cases and Controls by Gender, and Relationship to Index Case Odds ratios for various subgroups of first degree relatives, calculated with the Cox proportional hazards model, are presented in Tables 5.19 and 5.20. The first model tested (Table 5.19) includes four variables of interest: (i) age of onset of the index case, (ii) gender of the index case, (iii) relationship to the index case, and iv) gender of the relative. The odds ratios for age of onset and gender of the relative were both significant. The model yielded an odds ratio of 2.236 (95% C.I. = 1.310-3.818) for first degree relatives of early onset cases, and an odds ratio of 1.900 (95% C.I. = 1.265-2.856) for first degree relatives of late onset cases, in relation to first degree relatives of controls. Female first degree relatives had an odds ratio of 2.099 (95% C.I. = 1.383-3.185) in relation to male first degree relatives of all cases and controls. The odds ratios for relationship to the index case and gender of the index case were not significant. A second model (reduced model in Table 5.20) included only the two variables that reached significance in the first model: age of onset of the index case and gender of the relative. Again, this model yielded an odds ratio of 2.306 (95% C.I. = 1.358-3.917) for first degree relatives of early onset cases, and an odds ratio of 1.851 (95% C.I. = 1.233-2.779) compared to first degree relatives of controls, and an odds ratio of 2.081 (95% C.I. = 1.371-3.158) for female first degree relatives of all cases and controls compared to male first degree relatives. The difference between these two models (Table 5.19 and 5.20) is non-significant (LRS = 4.126 df = 2, p>0.05). 85 Table 5.19: Odds Ratios for Various Subgroups of First Degree Relatives of AD Cases and Controls Variable Odds Ratio 95% C.I. Age of Onset Controls 1.00 Late Onset 1.900 1.265-2.856 Early Onset 2.236 1.310-3.818 Gender of the Proband Female 1.00 Male 1.290 0.8905-1.869 Relationship to Proband Sibs 1.00 Parents 1.349 0.9077-2.005 Gender of the Relative Male 1.00 Female 2.099 1.383-3.185 Deviance = 1718.864 Likelihood Ratio Statistic on 5 df = 30.365 p<0.001 Table 5.20: Odds Ratios for Various Subgroups of First Degree Relatives of AD Cases and Controls (Reduced Model) Variable Odds Ratio 95% C.I. Age of Onset Controls 1.00 Late Onset 1.851 1.233-2.779 Early Onset 2.306 1.358-3.917 Gender of the Relative Male 1.00 Female 2.081 1.371-3.158 Deviance = 1722.990 Likelihood Ratio Statistic on 3 df = 26.239 p<0.001 86 5.10 Risk Estimates for Female and Male First Degree Relatives of All C a s e s With and Without Corrections for the Gender Ratio in the General Population Kaplan-Meier age-specific risks for female and male first degree relatives of all cases are presented in Table 5.21. The cumulative lifetime risk to age 88 years for female first degree relatives of cases (9.5 ± 1.8%) is significantly different from the cumulative lifetime risk to male first degree relatives of cases (4.6 ± 1.4%) ( Z = 2.15, p=0.01). Female first degree relatives have an earlier onset than male first degree relatives of cases, and the difference between the two risk curves is significant (see Figure 5.11)(Log Rank x 2 = 6.59, p<0.025). When the 1.2:1 correction was applied (see Figure 5.12), the cumulative lifetime risk estimates remained significantly different at age 88 years (Z=1.74, p=0.04), however the lifetime risk curves were no longer significantly different (Log Rank x 2 =2.64, p>0.05). When the 1.5:1 correction was made (see Figure 5.13), neither the cumulative lifetime risks (Z=1.14, p>0.05), nor the cumulative lifetime curves differed significantly (Log Rank X 2 =0.82, p>0.05). 87 Table 5.21: Age-Specif ic Risks for Female and Male First Degree Relatives of Index C a s e s Relative Age Risk (%) to Female First Risk (%) to Male First Degree Relatives of Degree Relatives of Probands ± 95% C.I. Probands ± 95% C.I. 65 0.0 0.0 66 0.1 (0.0-0.9) 0.0 67 0.1 (0.0-0.9) 0.1 (0.0-1.0) 68 0.2 (0.1-1.0) 0.1 (0.0-1.0) 69 0.5 (0.2-1.4) 0.1 (0.0-1.0) 70 0.7 (0.3-1.6) 0.3 (0.1-1.2) 71 1.0 (0.5-2.0) 0.5 (0.2-1.5) 72 1.3 (0.7-2.4) 0.5 (0.2-1.5) 73 1.4 (0.8-2.7) 0.5 (0.2-1.5) 74 1.6 (0.9-2.9) 0.9 (0.4-2.3) 75 2.4 (1.4-4.0) 1.2 (0.5-2.6) 76 2.8 (1.8-4.6) 1.2 (0.5-2.6) 77 3.1 (1.9-4.9) 1.2 (0.5-2.6) 78 3.8 (2.5-5.9) 2.2 (1.1-4.3) 79 4.6 (3.1-6.9) 2.6 (1.4-5.0) 80 5.5 (3.8-8.1) 2.6 (1.4-5.0) 81 5.5 (3.8-8.1) 3.2 (1.7-5.9) 82 6.3 (4.3-9.1) 3.2 (1.7-5.9) 83 7.1 (4.9-10.2) 3.9 (2.1-7.2) 84 7.6 (5.3-10.8) 4.6 (2.5-8.5) 85 8.7 (6.1-12.4) 4.6 (2.5-8.5) 86 8.7 (6.1-12.4) 4.6 (2.5-8.5) 87 9.5 (6.6-13.7) 4.6 (2.5-8.5) 88 9.5 (6.6-13.7) 4.6 (2.5-8.5) 88 Figure 5.11: Age-Specific Risk (%) to Female and Male First Degree Relatives of Cases 10 i n i o s c o o i O i - N n ^ o i D N o i o i O i - N n ^ i n i D s c o < o ( 0 ( 0 ( o < o i ^ N ' i ^ . r ^ i ^ r > . h < h « h > i ^ c o c o o o c o o o c o c o c o c o AGE (years) Females Males 89 Figure 5.12: Age-Specific Risk (%) to Female and Male First Degree Relatives of Cases Adjusted for a Gender Ratio (F:M) of 1.2:1 Figure 5.13: Age-Specific Risk (%) for Female and Male First Degree Relatives of Cases Adjusted for a gender ratio (F:M) of 1.5:1 6. Discussion It has been estimated that approximately 90% (Cruts et al., 1996) of all AD cases are sporadic; i.e., these cases are from families that do not seem to follow an autosomal dominant mode of inheritance. To date, to the best of my knowledge, no study has been published which looks at the familial risks of such sporadic AD cases specifically. It was possible to determine the effect of taking out all the Alzheimer's disease cases from families suspected to have an autosomal dominant mode of inheritance and their respective first degree relatives from all risk calculations by comparing the results of the present study with those of a previous study which included AD cases taken from the same clinic population (Hirst 1993). The previous study calculated survival data for 1867 first degree relatives of 338 early and late onset AD cases from the AD Clinic. These data were then compared with those for a small group of controls taken from the Canadian Study of Health and Aging (CSHA). The study sample consisted of all cases presenting at the AD Clinic with a diagnosis of probable AD or definite AD, who had sufficient family history information, including possible autosomal dominant cases, and clear familial AD cases. Figure 6.1 shows the risk curves for all first degree relatives from the previous study and for first degree relatives of strictly non-autosomal dominant cases from the present study. The difference between the two cumulative lifetime risk estimates to age 88 years (23.07± 2.74% vs 7.9± 1.5% , Z=4.85, p<0.001) was significant. The inclusion of autosomal dominant cases decreased the age of onset, and the risk curve of all first degree relatives increased at a greater rate than the one for first degree relatives of strictly non-autosomal dominant cases of AD, beginning at age 36 years and continuing to age 88. Inclusion of families with multiple cases of AD significantly increased the risk estimates for first degree relatives. This is not surprising, as the majority of all non-autosomal dominant AD cases did not have even one affected first degree relative. 92 Figure 6.1: Age-Specific Risk (%) to First Degree Relatives of All Cases Vs Non-Autosomal Dominant Cases 9 3 6.1 The Difference in Genetic Loading in Early Onset Non-Autosomal Dominant C a s e s and Late Onset Non-Autosomal Dominant Cases of Alzheimer's Disease The much publicized recent discovery of the two early onset autosomal dominant AD genes, presenilin-1 and presenilin-2 (Sherrington et al., 1995; Levy-Lahad et al., 1995), along with the knowledge of the existence of another early onset autosomal dominant AD gene, B APP (Goate et al., 1991), has suggested that early onset cases may be more genetically loaded than late onset AD cases. Along the same lines of thought, if AD is one disease, in which there is no real basis for the division into early and late onset cases, and for which the cause is polygenic (excluding all autosomal dominant cases for which an autosomal dominant gene has been found), one might expect that the early onset cases would be more genetically loaded, and therefore would express the disease phenotype at an earlier age, as would their first degree relatives in comparison with first degree relatives of late onset cases. Given these assumptions, it was interesting that results from the present study showed that the cumulative lifetime risk estimate to first degree relatives of early onset non-autosomal dominant cases of AD was not significantly greater or lower than the cumulative lifetime risk estimate to first degree relatives of late onset AD cases i.e. the risks did not differ significantly. Even more surprising was that the risk to first degree relatives of early onset cases did not increase at an earlier age than the risk to first degree relatives of late onset cases, as could be expected if a polygenic model was assumed. This result was clearly different from what was seen in the previous study by Hirst (1993) which included known and suspected autosomal dominant AD families. In the previous study, the risk to first degree relatives of early onset cases increased at a greater rate at an earlier age compared to the risk to first degree relatives of late onset cases. This probably resulted 94 from the 12 early onset autosomal dominant AD families included, as cases coming from the same FAD family do tend to have similar ages of onset (Bird et al., 1989). There was concern in regards to the possibility of bias in this study against finding and earlier age of onset for first degree relatives of early onset cases and a later onset for first degree relatives of late onset cases simply due to the ommision of cases with first degree relatives with a similar age of onset, as possible autosomal dominant AD families. However, this was kept in mind throughout the screening process, and all families in which an autosomal dominant pattern of inheritance could not be ruled out were excluded from our study. This resulted in 1.4% of early onset AD cases, and 1.7% of late onset AD cases being ommited as "possible" autosomal dominant AD cases. The difference between these two proportions (2/157 and 7/306) was not significant (p>0.05). Therefore I believe that the results of this study were not biased against finding certain trends in non-autosomal dominant AD familial aggregation, and do in fact reflect the true risk to first degree relatives of early and late onset non-autosomal dominant AD cases. In fact, Hirst (1993) found that although the risk to first degree relatives of early onset AD cases increased at an earlier age than the risk to first degree relatives of late onset AD cases, the risk curves for both groups showed a similar increase at later ages, with the lifetime risk curves not differing significantly from each other to age 95 years. The results of the "present" study refute the hypothesis that early onset non-autosomal dominant cases are more genetically loaded than late onset non-autosomal dominant cases of AD, forcing one to re-think the whole idea of dividing AD cases according to age of onset groups. In fact, when the data were analyzed with the Cox proportional hazards model, the odds ratio for first degree relatives of early onset cases in relation to the risk to first degree relatives of late onset cases was not significantly different. Therefore, it 95 might be more appropriate to look at all non-autosomal dominant AD cases as one group instead of dividing them by age of onset. 6.2 Genetic Loading in Early Onset and Late Onset Cases Compared to Controls A comparison between the cumulative lifetime risk estimates for first degree relatives of late onset cases and of controls up to age 88 years showed a significant difference between the two. The risk to first degree relatives of late onset cases also increased at a significantly greater rate as of age 59 to age 88 compared to the more gradual increase in risk throughout life to first degree relatives of controls. When the first degree relatives of early and late onset cases were compared to first degree relatives of controls in the Cox proportional hazards model, both groups of first degree relatives of AD cases had a significantly increased odds ratio in relation to first degree relatives of controls. These results indicated that both early and late onset cases have greater familial AD aggregation than found in the general population and therefore a genetic factor(s) may have a role in the cause of non-autosomal dominant AD. It would have been of great interest to be able to genotype a portion of individuals for apolipoprotein E status to determine if this was in fact a genetic factor at play in non-autosomal dominant cases of AD. Unfortunately, appropriate blood samples for such typing could not be obtained within the time frame of this study (This was a factor related to CSHA which was beyond my control). 6.3 Gender of the First Degree Relative The female gender is a known AD risk factor. It was therefore not surprising to find that the cumulative lifetime risk estimate to age 88 for female first degree relatives of both early onset AD cases and controls were significantly different than those for male first 96 degree relatives of the same groups. When first degree relatives of all AD index cases were grouped together, the resulting difference between the cumulative lifetime risk estimates to age 88 for female and male first degree relatives was significant. As seen in the control group, the risk to female first degree relatives of cases increased at a greater rate as of age 63, compared to the risk to male first degree relatives of cases. This increased risk for females remained up to age 88 years. At first glance, these results were greatly different from the conclusions of Hirst (1993), where it was reported that the cumulative lifetime risk estimates to female and male first degree relatives of cases to age 95 years were not significantly different from each other. However, if the difference of proportions test was calculated on the cumulative lifetime risk estimates up to age 88 years, the difference between the cumulative lifetime risk estimates to female and male first degree relatives of cases from the "previous" study became significant, which was in agreement with the results for the "present" study. A significant difference between the results of the "present" study and the "previous" study lies in the results reported for the control population. The "previous" study reported that both the cumulative lifetime risk estimates and the pattern of the lifetime risk curves for female and male first degree relatives of controls were not significantly different from each other. This was very different from the results of the "present" study, where both statistical tests showed statistically significant differences between the risk to female and male first degree relatives. In fact, the odds ratio for female first degree relatives in relation to male first degree relatives of controls was 2.22 (95% C.I .= 1.21-4.03). This difference between the two studies could be attributed to a doubling of the total number of first degree relatives of controls included in the "present" study compared to the "previous" study. (The previous study included only a limited number of CSHA controls which were available at that time) 97 In view of these results, it was of interest to determine if the increased risk to female first degree relatives of cases compared to male first degree relatives of cases was simply a reflection of the situation existing in the general population. Thus corrections were made for gender (F:M) ratios of 1.2:1 and 1.5:1, taken from the CSHA. The data showed that the 1.5:1 correction was sufficient to shift the cumulative lifetime risk of males to a level not significantly different from the cumulative lifetime risk to females. The lifetime risk curves were also not significantly different from each other after corrections. The result of comparisons between the risk to first degree relatives of cases and controls indicated that early and late onset non autosomal dominant AD cases have a greater familial aggregation, which may be due to greater genetic loading, than found in the general population. This genetic component might work irrespective of gender, as suggested by the results of the gender ratio correction. Therefore an increased odds ratio for female first degree relatives of cases in relation to male first degree relatives was observed, which was not at a magnitude greater than the one found in the general population. The genetic component might therefore balance out the risk ratio between females and males, as can be seen by the gender risk ratio found in first degree relatives of cases, which was close to non-significance. These results also argue against a gender-influenced AD gene. Such a gene, if present, would lead to the expectation that the risk to females would be greater than the risk to males even after correction for the gender ratio found in the general population. The CSHA data (CSHA 1994(a)) represents the most accurate data available on AD prevalence in the Canadian population. However, these data are only available for the population aged 65 and above. For this reason, the gender ratio corrections could only be applied to first degree relatives aged 65 and above. I believe that omitting the younger first degree relatives from the corrections did not result in a significant difference, as the risk to 98 female and male first degree relatives of both early and late onset cases were relatively similar before the age of 65 (see Figures 5.2 and 5.3) 6.4 Relationship The relationship (parent or sib of the index case) between the first degree relative and the index case did not appear to be a significant risk modifying factor. The cumulative lifetime risk estimate for parents and sibs were very similar in both late onset cases and in controls, as were the patterns of the lifetime risk curves. In the case of first degree relatives of early onset cases, the risk curve for sibs consisted of a single peak and subsequent plateau. Very little emphasis should be laid on this result, as the sample size of sibs of early onset cases was very small even at earlier ages (i.e. N=222 sibs of early onset cases surviving up to age 65 compared to N=761 sibs of late onset cases) and the number of surviving sibs of early onset cases continued to decrease at a drastic rate with increasing age (i.e. only 11 sibs of early onset cases survived up to age 83 years, while 93 sibs of late onset cases survived and remained disease free at this same age). Therefore, it is possible that we have not yet studied the sibs of early onset cases long enough to be able to make solid conclusions in regards to their risk. The finding that the cumulative lifetime risk estimates for sibs and parents of late onset cases and controls were not significantly different from each other up to age 88 years, as are the patterns of their respective risk curves suggested that there was no generational non-genetic risk factor (non-genetic component related to a specific cohort). However, we could not extend this conclusion to non generational environmental factors such as exposure to aluminum, smoking, or head trauma. 99 6.5 Gender of the Index Case The gender of the index case did not seem to contribute any important information to the cumulative lifetime risk estimates for the first degree relatives of cases and controls. In all three groups (early onset cases, late onset cases, and controls), the cumulative lifetime risk estimates calculated for first degree relatives of female index cases did not differ significantly from the cumulative lifetime risk estimates calculated for first degree relatives of male index cases. The pattern of the lifetime risk curves was also not significantly different in all three groups. These results are in agreement with the study of Korten et al., (1993) in which this variable was tested for first degree relatives of 91 AD cases. 6.6 The influence of Genetic Factors in Non-Autosomal Dominant Cases of Alzheimer's disease The results of the present study lead us to believe that there is a genetic factor at work, even in the strictly non-autosomal dominant cases of AD. The mode of action of such a genetic factor remains unclear. Two previously characterized gene polymorphisms could have a role in this AD population. There may also be as yet unidentified genes. The apolipoprotein E E4 allele has been shown in multiple studies to increase the risk of developing AD in heterozygous carriers (£4/ex), and even more so in homozygous carriers (E4/E4). A dosage effect has also been shown between the age of onset of AD and the number of E4 alleles, in both familial and sporadic late onset AD cases (Poirier 1993; Corder et al., 1993). However, this inverse dose-related effect is stronger in familial late onset AD cases (Lautenschager et al., 1997). Multiple studies have now looked for the same type of correlation between age of onset of AD and E4 allele dosage in early onset AD cases; however none has been able to find such a correlation in either sporadic or familial 100 early onset AD cases (Lannfelt et al., 1994; VanDuijn et al., 1994; Perez-Tur et al., 1995). In fact, one study has shown that the e4 allele is only significantly elevated in early onset AD cases with a family history of one or more first or second degree relative(s) with late onset AD (>60 years)(Perez-Tur et al., 1995). The segregation of such a genetic susceptibility allele in our case population would be in accordance with our results, as the risk to first degree relatives of early onset AD cases did not increase significantly earlier than the risk to first degree relatives of late onset AD cases. More recently, there have been suggestions that an intronic polymorphism in the PS-1 gene could also be associated with an increased risk of early and late onset AD. It is not quite clear as to how this polymorphism works to increase the risk of developing AD, but it is most probably in linkage disequilibrium with a mutation in the PS-1 gene or a gene situated close to it. There has been a reported sporadic case of AD which has one of the previously characterized PS-1 mutations found in autosomal dominant early onset AD families (Sandbrink et al., 1996). However, it is most likely that this is a rare occurrence, and it would not account for many non-autosomal dominant cases. Additionally, if this individual passes the mutation on to the next generation, she would in fact represent a new mutation. Her offspring would then have to be followed longitudinally. Although there is support for a genetic component to non-autosomal dominant AD cases, we cannot yet rule out that the increased familial aggregation in non-autosomal dominant AD families from this study is determined at least in part by a common familial non-genetic factor such as level of education, exposure to aluminium, or some yet unknown factor. In order to determine if such familial non-genetic factors do influence the familial aggregation of non-autosomal dominant AD, adoption studies would be required. Unfortunately, because AD is a late onset disease, the retrieval of medical information is 101 difficult even on biological first degree relatives in normal conditions. The retrieval of medical records of adopted away children or of biological first degree relatives of adopted out individuals would be near impossible. Much evidence remains for non genetic (environmental) factors to have a role in the cause of AD. A recent study of Japanese men who migrated to Hawaii (the Honolulu-Asia Aging Study or HAAS) has put some emphasis back on environmental and cultural risk factors for AD (White et al., 1996). The results showed that the prevalence of AD in these Japanese born males was increased from that observed in Japan to a level approaching that for Caucasians of European ancestry, while the prevalence of vascular dementia in these same men remains close to that seen in the native Japanese population. The study shows the importance of environmental and/or cultural risk factors in the development of AD, as the Asian population generally has a lower prevalence of AD than other ethnicities, and a higher prevalence of vascular dementia (VsD). The AD/VsD ratio in Japan is estimated at 0.8, while the same ratio in the United-States and Europe has been in the range of 2 (White et al., 1996; CSHA 1994(a)). The AD/VsD ratio reported for these Japanese born men who moved to Hawaii was 1.5, a ratio that is intermediate between that of American and Japanese populations. This result supports the existence of environmental risk factors in AD. The precise contribution of genetic factors, non-genetic factors, and interactions between the two remains to be determined in non-autosomal dominant AD. It is most likely, in my opinion, that non-autosomal AD is multifactorial, and therefore early onset cases can result from a greater amount of non genetic factors which will push the individual toward the threshold at an earlier age. This is in line with our findings that first degree relatives of early onset cases do not have an earlier onset than first degree relatives of late onset cases 102 (i.e. the non-genetic factor (e.g., head trauma) affecting the age of onset of the index case would not influence the age of onset of first degree relatives). Several lines of evidence now exist for the interaction of genetic and environmental risk factors in the development of AD. The first comes from a study by Mayeux et al., (1995) who showed that a history of head trauma increases the risk of developing AD only in individuals carrying the e4 allele of apoE. A second comes from a study in which it was shown that smoking decreases the risk of developing AD significantly only in individuals with a family history of dementia who carry the e4 allele (Van Duijn et al., 1995). The results of the present study do not allow us to differentiate between genetic and non-genetic factors that could have a role in increasing the familial aggregation in first degree relatives of cases compared to controls. In order to determine the contribution of genetic and non-genetic factors in non-autosomal dominant AD, twin studies are needed. Historically, twin studies in AD have been very small and the results were hard to interpret. This was caused by the relatively late onset age of AD. Because AD is a late onset disorder, it is difficult to gather a sample of twin pairs, in which both twins have survived to the age of increased AD susceptibility. Also, as AD is age dependent and variation in age of onset is common even within MZ twin pairs, twin pairs that are discordant at the time of initial study may become concordant within a few years. The age of the sample does affect the concordance rate (Breitner et al., 1993 & 1995; Bergem 1994). In order for a twin study of AD to yield significant results, a large population would have to be sampled, and longitudinal follow up would be required. This would of course entail great costs over a long period of time. However, a study of this kind would be beneficial in the search for the answers to the Alzheimer's disease puzzle. 6.7 Strengths and Weaknesses of our Study 103 The great advantage of our study lies in the size of both our case and control groups. The one exception (sibs of early onset cases) is discussed in section 6.4. This study represents one of the biggest survival analysis for first degree relatives of AD cases to date. Sample size is of great importance in the study of familial risks in AD. Alzheimer's disease is a late onset dementia, and therefore the number of first degree relatives that survive to or beyond the age of onset of the index case can become very small. This becomes very apparent in the eighth and ninth decades of life. Since the Kaplan-Meier method is a chain multiplication of conditional probabilities, having very small sample sizes at the final ages can lead to very great standard errors, which can foil results and greatly reduce the significance of findings. Although the CSHA elderly Canadian sample could not be used for comparisons with the early onset sample, there were many advantages to using this population as our control group. As mentioned earlier, the CSHA sample is a Canada wide random sample of the elderly population aged 65 and above. The screening of all sampled individuals with the 3MS test, and the use of a high cut off score of 80 points, gave us added assurance that we did not include any individuals showing mild cognitive decline in our control sample. The greatest source of error in our study remains the method of assigning an age of onset to first degree relatives for which we do not have documentation about age of onset. The use of the age at which autopsy or clinical assesment was performed (latest age possible) will result in an increase of the risk estimate, by pushing the onset of an individual at a late age, a time at which fewer individuals survive, thus resulting in an increased risk estimate. This resulting bias can only grow smaller as we continue to collect data on families of cases presenting at the clinic, as there is now a greater awareness of dementia, and therefore there is better documentation of such disorders for first degree relatives. 104 The ApoE e4 allele has been shown to increase the risk of early and late onset AD and decrease the age of AD onset in familial and sporadic late onset AD. Initially, ApoE genotyping was to be included as a part of this study, however due to unforeseen delays in the second phase of CSHA, Apo E genotyping was made impossible. The inclusion of ApoE genotyping in the present study would have permitted us to determine if familial aggregation is greater in families of ApoE e4 positive non autosomal dominant AD cases compared to ApoE e4 negative cases. Therefore it would have given us insight into whether ApoE is responsible for the increased familial aggregation observed in the non-autosomal dominant cases compared to controls. In one study (Farrer et al., 1995), in which they looked at the risk of AD in first degree relatives of AD probands genotyped for Apo E, first degree relatives of e4 carrying probands had an increased risk of AD compared to first degree relatives of ex/ex probands (x= 2 or 3). However, it was found that first degree relatives of e4/e4 probands did not have a significantly increased risk of AD, nor did they have a significantly earlier age of onset of AD than first degree relatives of e4/ex probands. In fact, the mean age of onset for first degree relatives of e3/e3, £4/ex, and e4/e4 probands were not significantly different from each other (Farrer et al., 1995). This study did not differentiate autosomal dominant AD cases from non-autosomal dominant AD cases, nor did they differentiate between early and late onset probands. 105 7. Summary The results of this study suggest that there is a genetic component to non-autosomal dominant cases of Alzheimer's disease. However, the genetic loading in early onset non-autosomal dominant cases does not seem to be strikingly greater than in late onset non-autosomal dominant AD cases, although small numbers of affected first degree relatives may have precluded the detection of more subtle differences. Our results show that in non-autosomal dominant AD cases, the division of cases by age of disease onset may not be a valid criterion for separation, as there was no indication of increased cumulative lifetime risk or decreased age of onset for first degree relatives of early onset cases in comparison with first degree relatives of late onset cases. This study also showed that the genetic loading in strictly non-autosomal dominant cases of AD is greater than in the general population. This genetic loading both increases the cumulative lifetime risk estimate and decreases the average age of onset for first degree relatives of non-autosomal dominant cases, in comparison to the cumulative lifetime risk estimate and average age of onset of first degree relatives of controls. The study did not find any evidence of increased risk to various sub-groups of first degree relatives of cases (i.e. sibs and parents, first degree relatives of female and male index cases, and female and male first degree relatives of index cases), which is not simply a reflection of the increased risk in the same subgroups in the general Canadian population. For example, the increased cumulative lifetime risk to female first degree relatives of cases in comparison to the cumulative lifetime risk to male first degree relatives of cases was shown to be a simple reflection of the increased risk to females in the general population. 106 The size of our samples of first degree relatives make this study one of the largest survival analyses to date, and therefore our results should be representative of the true situation occurring in non-autosomal dominant cases. The risk to first degree relatives of strictly non-autosomal dominant AD cases is significantly lower than previous estimates which included first degree relatives of all AD cases together (including possible autosomal dominant cases and clear FAD families) in their risk calculations. The risk estimates generated from this study will be useful in the counselling of family members of non-autosomal dominant cases of AD from the BC population, which are concerned about their own risks of developing the disease. These risk estimates, when used in counselling non-autosomal dominant AD cases and their families, will serve to reassure them that although their risk is higher than in the general population, it is nowhere near the previous reports of 50%. In fact, the risk to first degree relatives of non-autosomal dominant AD up to age 88 is only twice as high as the general population risk of 4.1%. Delays in the second phase of CSHA meant ApoE genotyping could not be included as a part of the present study. Future studies of non-autosomal dominant AD cases should be designed in order to incorporate ApoE genotyping as a covariate, as ApoE e4 has been estimated to account for up to 40-50% of AD cases and could very well be the risk enhancing factor in non-autosomal dominant AD cases. 107 Bibliography Aisen PS & Davis KL: Inflammatory Mechanisms in Alzheimer's Disease: Implications for Therapy. Am J Psychiatry 1994;151:1105-1113 Alzheimer's Disease Colloborative Group: Apolopoprotein E genotype and Alzheimer's disease. Lancet 1993; 342: 737-738 Alzheimer's Disease Colloborative Group: The Structure of the Presenilin 1 (S182) Gene and Identification of Six Novel Mutations in Early Onset AD Families. Nature Genetics 1995;11:219-222 American Psychiatric Association: Diagnostic and Statistical Manual of Mental Disorders, 3rd ed revised, American Psychiatric Association,. Washington, DC, 1987 American Psychiatric Association: Diagnostic and Statistical Manual of Mental Disorders, 4th ed, American Psychiatric Association, Washington, DC, 1994 Andersen K, Launer LJ, Ott A, et al.: Do nonsteroidal Anti-Inflammatory Drugs Decrease the Risk for Alzheimer's Disease? The Rotterdam Study. Neurology 1995;45: 1441-1445 Andreasen NO, Endicott J., Spitzer RL., Winokur G.: The family History Method Using Diagnostic Criteria. Reliability and Validity. Arch Gen Psychiatry 1977; 34: 1229-1235 Asada T, Kariya T, Yamagata Z, Kinoshita T, Asaka A: Apolipoprotein E Allele in Centenarians. Neurology 1996; 1484 Baldereschi M., Amato M.P., Nencini P., etal.,: Cross-National Interrater Agreement on the Clinical Diagnostic Criteria for Dementia. Neurology 1994; 44: 239-242 Barinaga M: Missing Alzheimer's Gene Found. Science 1995; 269: 917-918 Benjamin R, Leake A, McArthur FK et al.: Protective effect of apoE e2 in Alzheimer's Disease. Lancet 1994; 344: 473 Benwell MEM, Balfour DJK, Anderson JM: Evidence that Tobacco Smoking Increases the Density of (-)-[3H] Nicotine Binding Sites in Human Brain. J Neurochem 1988;50:1243-1247 Bierer LM., Silverman JM., Mohs RO, et al.,: Morbid Risk to First Degree Relatives of Neuropathologically Confirmed Cases of Alzheimer's Disease. Dementia 1992; 3: 134-139 Bird TD, Sumi SM, Nemens EJ, et al.: Phenotypic Heterogeneity in Familial Alzheimer's Disease: A Study of 24 Kindreds. Ann Neurol 1989; 25:12-25 108 Boissiere F, Pradier L, Delaere P, et al.: Regional and Cellular Presenilin 2 (STM2) Gene Expression in the Human Brain. NeuroReport 1996;7:2021-2025 Bonaiuto S, Rocca WA, Lippi A, et al.: Impact of Education and Occupation on the Prevalence of Alzheimer's Disease (AD) and Multi-lnfarct Dementia (MID) in Appignano, Macerata Province, Italy. Neurology 1990; 40(suppl 1): 346 Breitner JCS, Gau BA, Welsh KA, et al.: Inverse Association of Anti-Inflammatory Treatments and Alzheimer's Disease: Initial Results of a Co-twin Control Study. Neurology 1994;44:227-232 Breitner JCS, Magruder-Habib KM: Criteria for Onset Critically Influence the Estimation of Familial Risk in Alzheimer's Disease. Genetic Epidemiology 1989; 6: 663-669 Breitner JCS, Silverman JM, Mohs RC, Davis KL: Familial Aggregation in Alzheimer's Disease: Comparison of Risk among Relatives of Early- and Late-Onset Cases, and among Male and Female Relatives in Successive Generations. Neurology 1988; 38: 207-212 Brousseau T, Legrain S, Berr C et al.: Confirmation of the E4 allele of the apolipoprotein E gene as a risk factor for late-onset Alzheimer's disease. Neurology 1994; 44: 342-344 Burns A & Murphy D: Protection Against Alzheimer's Disease? Lancet 1996;348:420-421 Cai X-D, Golde TE, Younkin SG.: Release of Excess Amyloid B Protein from a Mutant Amyloid BProtein Precursor. Science 1993;259: 514-516 Campion D, Flaman J-M, Brice A, et al.,: Mutations of the Presenilin 1 Gene in Families with Early-Onset Alzheimer's Disease. Hum. Molec. Genet.1995, 4(12): 2373-2377 Canadian Study of Health and Aging Working Group (a): study methods and prevalence of dementia. Can Med Assoc J 1994; 150(6):899-913 Canadian Study of Health and Aging Working Group (b): Risk factors for Alzheimer's disease in Canada. Neurology 1994; 44: 2073-2080 Chartier-Harlin M-C, Crawford F, Houlden H et al.: Early-onset Alzheimer's disease caused by mutations at codon 717 of the p-amyloid precursor protein gene. Nature 1991;353: 844-846 Citron M, Oltersdorf T, Haass C, et al.: Mutation o the P-Amyloid Precursor Protein in Familial Alzheimer's Disease Increases p-Protein Production. Nature 1992; 360:672-674 Clements A, Walsh DM, Williams CH, Allsop WD: Effects of the Mutations Glu 2 2 to Gin and Ala 2 1 to Gly on the aggregation of a synthetic fragment of the Alzheimer's Amyloid P/A4 Peptide. Neuroscience Letters 1993;161:17-20 109 Cook DG, Sung JC, Golde TE, et al.: Expression and Analysis of Presenilin 1 in a Human Neuronal System: Localization in Cell Bodies and Dendrites. Proc Natl Acad Sci USA 1996;93: 9223-9228 Corder EH, Saunders AM, Risch NJ et al.: Protective effect of apolipoprotein E type 2 allele for late onset Alzheimer disease. Nat. Gen. 1994; 7:180-184 Corder EH, Saunders AM, Strittmatter WJ et al.: Gene Dose of Apolipoprotein E Type 4 Allele and the Risk of Alzheimer's Disease in Late Onset Families. Science 1993; 261: 921-923 Corder EH, Saunders AM, Strittmatter WJ et al.: Apolipoprotein E, Survival in Alzheimer's Disease Patients, and the competing Risks of Death and Alzheimer's Disease. Neurology 1995; 45:1323-1328 Crowley J, Breslow : Statistical Analysis of Survival Data. Ann Rev Public Health 1984; 5: 385-411 Cruts M, Hendriks L, Van Broeckhoven C: The Presenilin Genes: A New Gene Family Involved in Alzheimer Disease pathology. Hum Mol Genet 1996;5:1449-1455 Deng G, Su JH, Cotman CW: Gene Expression of Alzheimer-Associated Presenilin-2 in the Frontal Cortex of Alzheimer and Aged Control Brain. FEBS Letters 1996;394: 17-20 Dovey HF, Suomensaari-Chrysler S, Lieberburg I, Sinha S, Keim PS: Cells with a Familial Alzheimer's Disease mutation Produce Authentic B-Peptide. NeuroReport 1993; 4:1039-1042 Duff K, Eckman C, Zehr C, et al.: Increased Amyloid-B42(43) in Brains of Mice Expressing Mutant Presenilin 1. Nature 1996; 383: 710-713 EGRET State of the Art Epidemiological Computing. Reference Manual. Statistics and Epidemiology Research Corporation 1993. Farrer LA, Cupples LA, Van Duijn CM, et al.: Apolipoprotein E Genotype in Patients with Alzheimer's Disease: Implications for the Risk of Dementia Among Relatives. Ann Neurol 1995; 38: 797-808 Farrer LA, O'Sullivan DM, Cupples LA, et al.,: Assessment of Genetic Risk for Alzheimer's Disease among First-Degree Relatives. Ann Neurol 1989;25: 485-493 Felsenstein KM, Hunihan LW, Roberts SB: Altered Cleavage and Secretion of a Recombinant B-APP Bearing the Swedish Familial Alzheimer's Disease Mutation. Nature Genetics 1994;6: 251-256 Fisher LD & Van Belle G: Biostatistics. A Methodology for the Health Sciences John Wiley & Sons, Inc. New York. 1993 Forbes WF & McLachlan DRC: Furher Thoughts on the Aluminum-Alzheimer's disease link. J Epidemiol Community Health 1996;50:401-403 110 Forette F., Henry JF., Orgotozo JM., et al.,: Reliability of Clinical Criteria for the Diagnosis of Dementia. A Longitudinal Multicenter Study. Arch Neurol 1989; 46: 646-648 Forster DP, Newens AJ, Kay DWK, Edwardson JA: Risk Factors in Clinically Diagnosed Presinile Dementia of the Alzheimer Type: A Case-Control Study in Northern England. J Epidemiol Community Health 1995;49:253-258 Fratiglioni L, Ahlbom A, Viitanen M, Windblad B: Risk Factors for Late-Onset Alzheimer's Disease: A Population-Based, Case-Control Study. Ann Neurol 1993; 33: 258-266 Games D, Adams D, Alessandrini R, et al.: Alzheimer-Type Neuropathology in Transgenic Mice Overexpressing V717F B-Amyloid Precursor Protein. Nature 1995; 373: 523-527 Gedye A, Beattie BL, Tuokko H, Horton A, Korsarek E: Severe Head Injury Hastens Age of Onset of Alzheimer's Disease. JAGS 1989;37:970-973 Goate A, Chartier-Harlin M-C, Mullan M et al.: Segregation of a missence mutation in the amyloid precursor protein gene with familial Alzheimer's disease. Nature 1991; 349: 704-706 Golde TE, Estus S, Younkin LH, Selkoe DJ, Younkin SG: Processing of the Amyloid Protein Precursor to Potentially Amyloidogenic Derivatives. Science 1992;255: 728-730 Graves AB, Van Duijn CM, Chandra V, et al.: Alcohol and Tobacco Consumption as Risk Factors for Alzheimer's Disease: A Collaborative Re-Analysis of Case-Control Studies. International Journal of Epidemiology 1991;20 (suppl 2): S48-S57 Haass C, Hung AY, Schlossmacher MG, Teplow DB, Selkoe DJ: B-Amyloid Peptide and a 3-kDa Fragment Are Derived by Distinct Cellular Mechanisms. J Biol Chem 1993;268(5):3021-3024 Haass C, Hung AY, Selkoe DJ, Teplow DB: Mutations Associated with a Locus for Familial Alzheimer's Disease Result in Alternative Processing of Amyloid B-Protein Precursor. J Biol Chem 1994; 269(26): 17741 -17748 Haass C, Koo EH, Mellon A, Hung AY, Selkoe DJ (a): Targeting of Cell-Surface B-Amyloid Precursor Protein to Lysosomes: Alternative Processing into Amyloid-Bearing Fragments. Nature 1992;357:500-503 Haass C, Schlossmacher MG, Hung AY, et al.(b): Amyloid B-Peptide is Produced by Cultured Cells During Normal Metabolism. Nature 1992;359:322-325 Haass C & Selkoe DJ: Cellular Processing of B-Amyloid Precursor Protein and the Genesis of Amyloid B-Peptide. Cell 1993;75:1039-1042 Hardy J : Framing p-Amyloid. Nat. Genet. 1992;1: 133-234 111 Henderson VW, Paganini-Hill A, Emanuel CK, Dunn ME, Buckwalter JG: Estrogen Replacement Therapy in Older Women. Comparisons Between Alzheimer's Disease Cases and Nondemented Control Subjects. Arch Neurol 1994;51:896-900 Henderson VW, Watt L, Buckwalter JG: Cognitive Skills Associated with Estrogen Replacement in Women with Alzheimer's Disease. Psychoneuroendocrinology 1996;21(4):421-430 Hendrie HC, Hall KS, Hui S et al.: Apolipoprotein E Genotypes and Alzheimer's Disease in a Community Study of Elderly African Americans. Ann. Neurol. 1995; 37: 118-120 Hendriks L, Van Broeckhoven C: The BA4 Amyloid Precursor Protein Gene and Alzheimer's Disease. Eur J Biochem 1996; 237: 6-15 Hendriks L, Van Duijn CM, Cras P, et al.: Presinile Dementia and Cerebral Haemorrhage linked to a Mutation at Codon 692 of the B-Amyloid Precursor Protein Gene. Nature Genetics 1992 ;1:218-221 Higuchi S, Muramatsu T, Matsushita S, Arai H, Sasaki H: Presenilin-1 Polymorphism and Alzheimer's Disease. Lancet 1996;347:1186 Hirst CG: Empiric Risk to First-Degree Relatives of Alzheimer Disease Patients. Master's Thesis 1993 Hirst C, Yee IML, Sadovnick AD: Familial Risks for Alzheimer Disease From a Population-Based Series. Genetic Epidemiology 1994;11: 365-374 Hong CJ, Liu TY, Liu HC, et al.: e4 Allele of Apolipoprotein E Increases Risk of Alzheimer's Disease in a Chinese Population. Neurology 1996; 46: 1749-1751 Hsiao K, Chapman P, Nilsen S, et al.: Correlative Memory Deficits, AB Elevation, and Amyloid Plaques in Transgenic Mice. Science 1996; 274: 99-102 Huff FJ, Auerbach J, Chakravarti A, Boiler F: Risk of Dementia in Relatives of Patients with Alzheimer's Disease. Neurology 1988; 38: 786-790 Ignatius MJ, Gebicke-Harter PJ, Skene JHP et al.: Expression of apolipoprotein E during nerve degeneration and regeneration. Proc. Natl. Acad. Sci. 1986; 83:1125-1129 Iwatsubo T, Saido TC, Mann DM, Lee VM, Trojanowski JQ: Full-Length Amyloid Beta (1-42(43)) and Amino-Terminally modified and Truncated Amyloid Beta 42(43) Deposit in Diffuse Plaques. Am J Pathol 1996; 149(6): 1823-1830 Jarette JT &. Lansbury PT: Seeding "One-Dimensional Crystallization" of Amyloid: A Pathogenic Mechanism in Alzheimer's Disease and Scrapie? Cell 1993; 73: 1055-1058 112 Johansson K, Pedersen N, Berg S, et al.: Genetic and Environmental Factors in the Etiology of Alzheimer's Disease and Related Disorders: The Study of Dementia in Swedish Twins. Neurobiology of Aging 1992; 13 (Suppl 1): S68 Johnston JA, Froelich S, Lannfelt L, Cowburn RF: Quantification of Presenilin-1 mRNA in Alzheimer's Disease Brains. FEBS Letters 1996;394:279-284 Jordan BD, Baker FM, Barclay L, Rocca WA, Schoenberg BS: Head Trauma and Participation in Contact Sports as Risk Factors for Alzheimer's Disease. Neurology 1990;40(suppl 1):347 Kamino K, Sato S, Sakaki Y, et al.: Three Different Mutations of Presenilin 1 Gene in Early-Onset Alzheimer's Disease Families. Neuroscience Letters 1996;208:195-198 Kaplan EL, Meier PJ: Non-Parametric Estimations from Incomplete Observations. JASA 1996;53:457-481 Kawamata J, Tanaka S, Shimohama S et al.: Apolipoprotein E polymorphism in Japanese patients with Alzheimer's disease or vascular dementia. J Neurol Neurosurg Psychiatry 1994; 57:1414-1416 Kehoe P, Williams J, Lovestone S, et al.: Presenilin-1 Polymorphism and Alzheimer's Disease. Lancet 1996;347:1185 Khoury JE, Hickman SE, Thomas CA, et al.: Scavenger Receptor-Mediated Adhesion of Microglia to p-Amyloid Fibrils. Nature 1996; 382: 716-719 Korten AE, Jorm AF, Henderson AS, et al.,: Assessing the Risk of Alzheimer's Disease in First-Degree Relatives of Alzheimer's Disease Cases. Psychological Medicine 1993; 23: 915-923 Kosunen O., Soininen H., Paljarvi L., et al.,: Diagnostic Accuracy of Alzheimer's Disease: a Neuropathological Study. Acta Neuropathol 1996; 91:185-193 Kovacs DM, Fausett HJ, Page KJ, et al.: Alzheimer-Associated Presenilins 1 and 2: Neuronal Expression in Brain and Localization to Intracellular Membranes in Mammalian Cells. Nature Medicine 1996;2(2):224-229 Kuusisto J, Koivisto K, Kervinen K et al.: Association of apolipoprotein E phenotypes with late onset Alzheimer's disease: population based study. BMJ 1994; 309: 636-638 Landen M, Thorsell A, Wallin A, Blennow K: The Apolipoprotein E Allele e4 Does Not Correlate with the Number of Senile Plaques or Neurofibrillary Tangles in Patients with Alzheimer's Disease. J Neurol Neurosurg Psychiatry 1996; 61: 352-356 Lannfelt L, Lilius L, Nastase M et al.: Lack of association between apolipoprotein E allele e4 and sporadic Alzheimer's disease. Neuroscience Letters 1994; 169:175-178 113 Lautenschlager NT, Muller U, Kurz AF: Effects of Apolipoprotein E e4 Allele on Age at Onset May Be Limited to Late-Onset Familial Alzheimer's Disease. Neurology 1997; 48(suppl.): A428 Lee V.M.-Y, Balin BJ., Otvos L, Trojanowski JQ.: A68: A Major Subunit of Paired Helical Filaments and Derivatized Forms of Normal Tau. Science 1991; 251: 675-678 Lemere CA, Lopera F, Kosik KS, et al.: The E280A Presenilin 1 Alzheimer Mutation Produces Increased AB42 Deposition and Severe Cerebellar Pathology. Nature Medicine 1996; 2(10): 1146-1150 Levy-Lahad E, Wasco W, Poorkaj P, et al.(a): Candidate Gene for the Chromosome 1 Familial Alzheimer's Disease Locus. Science 1995; 269 973-977 Levy-Lahad E, Wijsman EM, Nemens E, et al.(b): A Familial Alzheimer's Disease Locus on Chromosome 1. Science 1995; 269:970-973 Li G, Shen YC, Li YT, et al.: A Case-Control Study of Alzheimer's Disease in China. Neurology 1992; 42: 1481-1488 Li G, Silverman JM, Smith CJ et al.,: Age at Onset and Familial Risk in Alzheimer's Disease. Am J Psychiatry 1995; 152: 424-430 Maestre G, Ottman R, Stern Y et al.: Apolipoprotein E and Alzheimer's Disease: Ethnic Variation in Genotypic Risks. Ann. Neurol. 1995; 37(2): 254-259 Mahley RW: Apolipoprotein E: Cholesterol Transport Protein with Expanding Role in Cell Biology. Science 1988; 240: 622-630 Mann DMA, Iwatsubo T, Cairns NJ, Lantos PL, et al.: Amyloid B Protein (AB) Deposition in Chromosome 14-Linked Alzheimer's Disease: Predominance of AB43(43). Ann Neurol 1996;40:149-156 Martins RN, Clarnette R, Fisher C et al.: ApoE genotypes in Australia: roles in early and late-onset Alzheimer's disease and Down's syndrome. Neuroreport 1995; 6: 1513-1516 Martyn CN, Osmond C, Edwardson JA, et al.: Geographical Relation Between Alzheimer's Disease and Aluminium in Drinking Water. Lancet 1989;59-62 Maruyama K, Tomita T, Shinozaki K, et al.: Familial Alzheimer's Disease-Linked Mutations at Val 7 1 7 of Amyloid Precursor Protein are Specific for the Increased Secretion of AB42(43). Biochem Biophys Res Comm 1996;227:730-735 Masters CL, Simms G, Weinman NA, et al.: Amyloid Plaque Core Protein in Alzheimer Disease and Down Syndrome. Proc Natl. Acad. Sci. USA 1985;82:4245-4249 Matsuo ES, Shin R-W, Billingsley ML et al.: Biopsy-Derived Adult Human Brain Tau Is Phosphorylated at Many of the Same Sites as Alzheimer's Disease Paired Helical Filament Tau. Neuron 1994 (13)989-1002 114 Mayeux R, Ottoman R, Maestre G, et al.: Synergistic Effects of Traumatic Head Injury and Apolipoprotein-e4 in Patients with Alzheimer's disease. Neurology 1995;45:555-557 Mayeux R, Ottoman R, Tang M-X, et al.: Genetic Susceptibility and Head Injury as Risk Factors for Alzheimer's Disease Among Community-Dwelling Elderly Persons and their First-Degree Relatives. Ann Neurol 1993;33:494-501 Mayeux R, Stern Y, Ottman R et al.: The Apolipoprotein e4 Allele in Patients With Alzheimer's Disease. Am. Neuro. Assoc. 1993; 34(5): 752-754 McGeer P., Rogers J.: Anti-Inflammatory Agents as a Therapeutic Approach to Alzheimer's Disease. Neurology 1992; 42: 447-449 McKhann G., Drachman D., Folstein M., et al.,: Clinical Diagnosis of Alzheimer's disease: Report of the NINCDS-ADRDA Work Group Under the Auspices of Department of Health and Human Services Task Force on Alzheimer's Disease. Neurology 1984;34: 939-944 Mortimer JA, French LR, Hutton JT, Schuman LM: Head Injury as a Risk Factor for Alzheimer's Disease. Neurology 1985;35:264-267 Moussaoui S, Czech C, Pradier L, et al.: Immunohistochemical Analysis of Presenilin-1 Expression in the Mouse Brain. FEBS Letters 1996;383:219-222 Mullan M, Crawford F, Axelman K, et al.: A Pathogenic Mutation for Probable Alzheimer's Disease in the APP Gene at the N-Terminus of B-Amyloid. Nature Genetics 1992; 1: 345-347 Muller HW, Gebicke-Harter PJ, Hangen DH and Shooter EM: A Specific 37,000-Dalton Protein that Accumulates in Regenerating but Not in Nonregenerating Mammalian Nerves. Science 1885; 228: 499-501 Murrell J, Farlow M, Ghetti B, Benson MD: A Mutation in the Amyloid Precursor Protein Associated with Hereditary Alzheimer's Disease. Science 1991;254:97-99 Myllykangas-Luosujarvi R & Isomaki H: Alzheimer's Disease and Rheumatoid Arthritis. British J of Rheumatology 1994;33(5):501 -502 Nagy ZS, Esiri MM, Jobst KA et al.: Influence of the Apolipoprotein E Genotype on Amyloid Deposition and Neurofibrillary Tangle Formation in Alzheimer's Disease. Neuroscience 1995; 69 (3): 757-761 Nalbantoglu J, Gilfix BM, Bertrand P. et al.: Predictive Value of Apolipoprotein E Genotyping in Alzheimer's Disease: Results of an autopsy Series and an analysis of Several Combined Studies. Ann Neurol 1994:36:889-895 Namba Y, Tomonaga M, Kawasaki H et al.: Apolipoprotein E immunoreactivity in cerebral amyloid deposits and neurofibrillary tangles in Alzheimer's disease and kuru plaque amyloid in Creutzfeldt-Jakob disease. Brain Research 1991; 541: 163-166 115 Nishimoto I, Okamoto T, Matsuura Y, et al.: Alzheimer Amyloid Protein Precursor Complexes with Brain GTP-Binding Protein G 0 . Nature 1993; 362: 75-79 Nishiyama K, Murayama S, Suzuki T, et al.: Presenilin 1mRNA Expression in Hippocampi of Sporadic Alzheimer's Disease patients. Neuroscience Research 1996;26:75-78 Nogushi S, Yamada N: Apolipoprotein E genotype and Alzheimer's disease. Lancet 1993; 342: 737-738 Okuizumi K, Onodera O, Tanaka H et al.: ApoE-e4 and early-onset Alzheimer's. Nat. Gen. 1994;7:10-11 Paganini-Hill A & Henderson VW: Estrogen Replacement Therapy and Risk of Alzheimer Disease. Arch Intern Med 1996;156:2213-2217 Payami H, Montee K, Grimslid H, Shattuc S, Kaye J: Increased Risk of Familial Late-Onset Alzheimer's Disease in Women. Neurology 1996; 46:126-129 Payami H, Montee K, Kaye JA et al.: Alzheimer's Disease, Apolipoprotein E4 and Gender. JAMA 1994; 271(17): 1316-1317 Payami H, Waggoner D, Litt M et al.: Genetics of Very-Late-Onset Alzheimer's Disease: A Prospective Study. Neurology 1995; 45(Suppl 4): A217 Peacock ML, Fink JK: ApoE e4 allelic association with Alzheimer's disease: Independent conformation using denaturing gradient gel electrophoresis. Neurology 1994; 44: 339-341 Perez-Tur J, Campion D, Martinez M, et al.: Evidence for Apolipoprotein E zA Association in Early-Onset Alzheimer's Patients With Late-Onset Relatives. Am J Med Genet 1995;60:550-553 Perez-Tur J, Wavrant-De Vrieze F, Lambert JC, Chartier-Harlin M-C, Alzheimer's Study Group: Presenilin-1 Polymorphism and Alzheimer's Disease. Lancet 1996;347:1560-1561 Plassman BL, Welsh KA, Brandt J, Page WF, Breitner JCS: Intelligence and Education as Predictors of Cognitive State in Late Life: A 50-year Follow Up. Neurology 1995;45:1446-1450 Poduslo SE, Herring K, Neal M: A Presenilin 1 Mutation in an Early Onset Alzheimer's Family: No Association with Presenilin 2. NeuroReport 1996;7:2018-2020 Poirier J, Davignon J, Bouthillier D et al.: Apolipoprotein E polymorphism and Alzheimer's disease. Lancet 1993; 342: 697-699 Polvikoski T, Sulkava R, Haltia M, et al.: Apolipoprotein E, Dementia, and Cortical Deposition of B-Amyloid Protein. N Engl J Med 1995; 333: 1242-1247 116 Ponte P, Gonzalez-DeWhitt P, Schilling J, et al.: A New A4 Amyloid mRNA Contains a Domain Homologous to Serine Proteinase Inhibitors. Nature 1988;331:525-527 Prince M, Cullen M and Mann A: Risk factors for Alzheimer's disease and dementia: A case-control study based on the MRC elderly hypertension trial. Neurology 1994; 44: 97-104 Raiha I, Kaprio J, Koskenvuo M, Rajala T, Sourander L: Alzheimer's Disease in Finnish Twins. Lancet 1996; 347: 573-578 Rebeck GW, Reiter JS, Strickland DK et al.: Apolipoprotein E in Sporadic Alzheimer's disease: Allelic Variation and Receptor Interactions. Neuron 1993; 11: 575-580 Rebeck GW, Perls TT, West HL et al.: Reduced apolipoprotein E4 allele frequency in the oldest old Alzheimer's patients and cognitively normal individuals. Neurology 1994;44:1513-1516 Rich JB, Rasmusson DX, Folstein MF et al.: Nonsteroidal Anti-Inflammatory Drugs in Alzheimer's Disease. Neurology 1995;45:51-55 Rogaev El, Sherrington R, Rogaeva EA, et al.: Familial Alzheimer's Disease in Kindreds with Missense Mutations in a Gene on Chromosome 1 Related to the Alzheimer's Disease Type 3 Gene. Nature 1995;376:775-778 Rogers J, Kirby LC, Hempelman SR, et al.: Clinical Trial of Indomethacin in Alzheimer's Disease. Neuroloy 1993;43:1609-1611 Rogers J, Webster S, Lue L-F, et al.: Inflammation and Alzheimer's Disease Pathogenesis. Neurobiology of Aging 1996;17(5):681-686 Rosenberg RN, Richter RW, Risser RC, et al.: Genetic Factors for the Development of Alzheimer Disease in the Cherokee Indian. Arch Neurol 1996; 53: 997-1000 Roses AD, Strittmatter WJ, Pericak-Vance MA et al.: Clinical application of apolipoprotein E genotyping to Alzheimer's disease. Lancet 1994 (342)1564-65 Roses AD, Saunders AM, Alberts MA et al.: Apolipoprotein E E4 Allele and Risk of Dementia. JAMA 1995; 273(5): 374 Rossor MN, Fox NC, Beck J, Campbell TC, Collinge J: Incomplete Penetrance of Familial Alzheimer's Disease in a Pedigree with a Novel Presenilin-1 Gene Mutation. Lancet 1996;347:1560 Sadovnick AD, MacDougall A, Burgess M, Karlinsky H: Genetic Screening and Counselling for Alzheimer's Disease: Predictive Testing and Screening for Risk Factors in Asymptomatic Individuals. (Submitted) 1997 Sadovnick AD, Irwin ME, Baird PA, Beattie BL: Genetic Studies on an Alzheimer Clinic Population. Genetic Epidemiology 1989; 6: 633-643 117 Sakoda S, Kuriyama M, Osame M et al.: Apolipoprotein E e4. Neurology 1994; 44:2420 Salib E & Hillier V: A Case-Control Study of Alzheimer's Disease and Aluminium Occupation. British Journal of Psychiatry 1996;168:244-249 Sandbrink R, Zhang D, Schaeffer S, et al.: Missense Mutations of the PS-1/S182 Gene in German Early-Onset Alzheimer's Disease Patients. Ann Neurol 1996; 40(2): 265-266 Saunders AM, Strittmatter WJ, Schmechel D et al. (b): Association of apolipoprotein E allele e4 with late-onset familial and sporadic Alzheimer's disease. Neurology 1993; 43: 1467-1472 Saunders AM, Schmader K, Breitner JCS et al. (b): Apolipoprotein E e4 allele distributions in late-onset Alzheimer's disease and in other amyloid-forming diseases. Lancet 1993; 342: 710-711 Schellenberg GD, Payami H, Wijsman EM, et al.: Chromosome 14 and Late-Onset Familial Alzheimer Disease (FAD). Am J Hum Genet 1993;53:619-628 Scheuner D, Eckman C, Jensen M, et al.: Secreted Amyloid B-Protein Similar to that in the Senile Plaques of Alzheimer's Disease is Increased in vivo by the Presenilin 1 and 2 and APP Mutations Linked to Familial Alzheimer's Disease. Nature Medicine 1996; 2(8): 864-870 Schmechel DE, Saunders AM, Strittmatter WJ et al.: Increased Amyloid p-peptide Deposition in Cerebral Cortex as a Consequence of Apolipoprotein E Genotype in Late-Onset Alzheimer Disease. Proc. Natl. Acad. Sci. 1993; 90: 9649-9653 Schneider LS, Farlow MR, Henderson VW, Pogoda JM: Effects of Estrogen Replacement Therapy on Response to Tacrine in Patients with Alzheimer's Disease. Neurology 1996;46:1580-1584 Scott WK, Growdon JH, Roses AD, Haines JL, Pericak-Vance MA (a): Presenilin-1 Polymorphism and Alzheimer's Disease. Lancet 1996;347:1186-1187 Scott WK, Roses AD, Haines, Pericak-Vance (b): Presenilin-1 Polymorphism and Alzheimer's Disease. Lancet 1996;347:1560 Seubert P, Oltersdorf T, Lee MG, et al.: Secretion of p-Amyloid Precursor Protein Cleaved at the Amino Terminus of the p-Amyloid Peptide. Nature 1993;361:260-263 Seubert P, Vigo-Pelfrey C, Esh F, et al.: Isolation and Quantification of Soluble Alzheimer's P-Peptide from Biological Fluids. Nature 1992;359:325-327 Shafer DR, & Nett LM: Medical Consequences of Cigarette Smoking: More Reasons to Quit. Seminars in Respiratory and Critical Care Medicine 1995; 16(2): 84-91 118 Sherrington R, Froelich S, Sorbi S, et al.: Alzheimer's Disease Associated with Mutations in Presenilin 2 is Rare and Variably Penetrant. Hum Mol Genet 1996;5(7): 985-988 Sherrington R, Rogaev El, Liang Y et al.: Cloning of a gene bearing missense mutations in early-onset familial Alzheimer's disease. Nature 1995; 375: 754-760 Silverman JH., Breitner JCS., Mohs R C , Davis KL.: Reliability of the Family History Method in Genetic Studies of Alzheimer's Disease and Related Dementias. Am J Psychiatry 1986; 143:1279-1282 Silverman JM, Li G, Zaccario ML, et al.,: Patterns of Risk in First-Degree Relatives of Patients With Alzheimer's Disease. Arch Gen Psychiatry 1994; 51:577-586 Silverman JM, Raiford K, Edland S, et al.,: The Consortium to Establish a Registry for Alzheimer's Disease (CERAD). Part VI. Family History Assessment: A Multicenter Study of First-Degree Relatives of Alzheimer's Disease Probands and Nondemented Spouse Controls. Neurology 1994; 44:1253-1259 Skene JHP, Shooter EM: Denervated sheath cells secrete a new protein after nerve injury. Proc. Natl. Acad. Sci. 1983; 80: 4169-4173 Smale G, Nichols NR, Brady DR, Finch CE, Horton WE: Evidence for Apoptotic Cell Death in Alzheimer's Disease. Experimental Neurology 1995; 133: 225-230 Smith AD, Johnston C, Sim E et al.: Protective effect of apoE e2 in Alzheimer's disease. Lancet 1994; 344: 473-474 Sobel E, Louhija J, Sulkava R et al.: Lack of association of apolipoprotein E allele E4 with late-onset Alzheimer's disease among Finnish centenarians. Neurology 1995; 45: 903-907 Sorbi S, Nacmias B, Forleo P, et al.: Missense Mutation of S182 Gene in Italian Families with Early-Onset Alzheimer's Disease. Lancet 1995;346:439-440 Stern Y, Gurland B, Tatemichi TK, et al.: Influence of Education and Occupation on the Incidence of Alzheimer's Disease. JAMA 1994;271:1004-1010 Strittmatter WJ, Saunders AM, Schmechel D et al.: Apolipoprotein E: High-avidity binding to p-amyloid and increased frequency of type 4 allele in late-onset familial Alzheimer's disease. Proc. Natl. Acad. Sci. 1993; 90:1977-1981 Strittmatter WJ, Weisgraber KH, Huang DY et al.: Binding of human apolipoprotein E to synthetic amyloid p peptide: Isoform-specific effects and implications for late-onset Alzheimer disease. Proc. Natl. Acad. Sci. 1993; 90: 8098-8102 Strittmatter WJ, Weisgraber KH, Goedert M et al.: Hypothesis: Microtubule Instability and Paired Helical Filament Formation in Alzheimer Disease Brain Are Related to Apolipoprotein E Genotype. Exp. Neurol. 1994; 125: 163-171 119 Su JH, Anderson AJ, Cummings BJ, Cotman CW: Immunohistochemical Evidence for Apoptosis in Alzheimer's Disease. NeuroReport 1994; 5: 2529-2533 Sullivan P, Petitti D, Barbaccia J: Head Trauma and Age of Onset of Dementia of the Alzheimer Type. JAMA 1987;257:2289-2290 Suzuki N, Cheung TT, Cai X-D, et al.: An Increased Percentage of Long Amyloid B Protein Secreted by Familial Amyloid B Protein Precursor (BAPP717) Mutants. Science 1994; 264:1336-1340 Suzuki T, Nishiyama K, Murayama S, et al.: Regional and Cellular Presenilin 1 Gene Expression in Human and Rat Tissues. Biochem Biophys Res Comm 1996;219: 708-713 Takashima A, Sato M, Mercken M, et al.: Localization of Alzheimer-Associated Presenilin 1 in Transfected COS-7 Cells. Biochem Biophys Res Comm 1996;227:423-426 Talbot C, Lendon C, Craddock N et al.: Protection against Alzheimer's disease with apoE e2. Lancet 1994; 343: 1432-1433 Tang M-X, Jacobs D, Stern Y, et al.: Effects of Oestrogen During Menopause on Risk and Age at Onset of Alzheimer's Disease. Lancet 1996;348:429-432 Taylor GA, Newens AJ, Edwardson JA, Kay DWK, Forster DP: Alzheimer's Disease and the Relationship Between Silicon and Aluminium in Water Supplies in Northern England. J Epidemiol Community Health 1995;49:323-324 Teng EL, & Chui HC: The Modified Mini-Mental State (3MS) Examination. J Clin Psychiatry 1987; 48:314-318 Tierney M O , Fisher RH., Lewis AJ., et al.,: The NINCDS-ADRDA Work Group Criteria for the Clinical Diagnosis of Probable Alzheimer's Disease: A Clinicopathologic Study of 57 Cases. Neurology 1988; 38: 359-364 Van Broeckhoven C: Presenilins and Alzheimer's Disease. Nature Genetics 1995;11: 230-232 Van Broeckhoven C, Backhovens H, Cruts M: APOE genotype does not modulate age of onset in families with chromosome 14 encoded Alzheimer's disease. Neuroscience Letters 1994; 169: 179-180 Van Duijn CM: Epidemiology of the Dementias: Recent Developments and New Approaches. Journal of Neurology, Neurosurgery, and Psychiatry 1996; 60: 478-488 Van Duijn CM, Clayton DG, Chandra V et al. (a): Interaction Between Genetic and Environmental Risk Factors for Alzheimer's Disease: A Reanalysis of Case-Control Studies. Genetic Epidemiology 1994; 11: 539-551 120 Van Duijn CM, Havekes LM, Van Broeckhoven C, et al.: Apolipoprotein E Genotype and Association Between Smoking and Early Onset Alzheimer's Disease. BMJ 1995;310:627-631 Van Duijn CM, de Knijff P, Cruts M et al. (b): Apolipoprotein E4 allele in a population-based study of early-onset Alzheimer's disease. Nat. Gen. 1994; 7: 74-78 Walden CC & Hegele RA: Apolipoprotein E in Hyperlipidemia. Ann Intern Med 1994; 120: 1026-1036 Wasco W, Pettingell WP, Jondro PD, et al.: Familial Alzheimer's Chromosome 14 Mutations. Nature Medicine 1995;1:848 West HL, Rebeck GW, Perls T et al.: The Prevalence of Apolipoprotein E4 in the Oldest Old:Connections With Alzheimer's Disease. Neurology 1994; 44(Suppl 2): A325 White LR, Petrovitch H, Ross W, et al.,: Prevalence of Dementia in Older Japanese-American Men in Hawaii. The Honolulu-Asia Aging Study. JAMA 1996; 276(12):955-960 Wisniewski T, Castano EM, Golabek A, Vogel T, Frangione B: Acceleration of Alzheimer's Fibril Formation by Apolipoprotein E In Vitro. Am J Pathol 1994; 145(5): 1030-1035 Wisniewski T, Palha J, Ghiso J, Frangione B: S182 Protein in Alzheimer's Disease Neuritic Plaques. Lancet 1995;346: Wragg M, Hutton M, Talbot C, Alzheimer's Disease Collaborative Group: Genetic Association Between Intronic Polymorphism in Presenilin-1 Gene and Late-Onset Alzheimer's Disease. Lancet 1996;347:509-512 Yamatsuji T, Matsui T, Okamoto T, et al.: G Protein-Mediated Neuronal DNA Fragmentation Induced by Familial Alzheimer's Disease- Associated Mutants of APP. Science 1996; 272: 1349-1352 Yan SD, Chen X, Fu J, et al.: RAGE and Amyloid-B Peptide Neurotoxicity in Alzheimer's Disease. Nature 1996: 382: 685-691 Yasuda M, Maeda K, Shimada K-l, et al.: Apolipoprotein E e4 Allele and Gender Difference in Risk of Alzheimer's Disease. Alzheimer's Research 1995; 1(2): 77-81 Yoshioka K, Miki T, Katsuya T, Ogihara T, Sakaki Y: The 717Val-->lle Substitution in Amyloid Precursor Protein is Associated with Familial Alzheimer's Disease Regardless of Ethnic Groups. Biochem Biophy Res Comm 1991; 178(3): 1141-1146 121 Appendix A Dementia Questionnaire 122 Proband Date Informant Interviewer Relative Age (if living) Age (at death) Cause of death Dementia Questionnaire MEMORY Did (does) (the subject) have Dont know Date any problems with 1) memory 2) remembering people's names 3) recognizing familiar faces 4) finding way about indoors 5) finding way on familiar streets 6) remembering a short list of items 7) did trouble with memory begin suddenly or slowly 8) Has the course of the memory problems been a steady downhill progression or have there been abrupt declines 9) Ever see a doctor for memory problems 10) If yes, what was the cause given? EXPRESSION 11) Ever have trouble finding the right word or expressing self 12) talking become less over time 13) tendency to dwell in the past DAILY FUNCTIONING 14) trouble with household tasks 15) handling money 16) grasping situations or explanations 123 (DAILY FUNCTIONING CONT'D) YES NO Don't know Date 17* difficulty at work (check If N/A 18) trouble dressing or caring for self 19) trouble feeding self 20) controlling bladder and bowels 21) agitation and nervousness OTHER PEOBLEMS 22) High blood pressure 23) Stroke 24) more than 1 stroke 25) is one side of the body weaker than other side — 26) Parkinson's disease (tremors, shuffling gait, rigidity of limbs _ 27) Injury to the head resulting in a loss of consciousness for more than a second or two _ 28) Seizure or fits _ 29) Syphilis . _ 30) Diabetes _ 31) drinking problem (if alcoholism suspected explore further SADSSxs) 32) did memory problems coincide with drinking 33) ever depressed or sad for two weeks' or more (if Depression suspected explore further SADSSxs) 34) If yes, ever seek treatment 35) ever very high, euphoric, top of the world 36) If yes, ever seek treatment 124 (OTHER PROBLEMS CONT'D) YES NO Dont know Date 37) ever seek psychiatric or psychological help for any reason 38) If yes, ever hospitalized for psychiatric Illness Where? , 39) Down's syndrome 40) other medical problems we have not talked about ; MEDICAL CONTACTS 42) Name and address of first doctor seen for problems 43) ever receive medications . 44) a neurological or psychiatric exam 45! CAT scan 46) ever in an institution Where? 47) What was diagnosis given for problems ' RECOGNITION OF PROBLEM 48) Who was the first person to notice something wrong 49) What was noticed 50) When was the last time (the subject) seemed to be really welL or his old self? -125 Appendix B Autosomal Dominant Pedigrees Individual without dementia Individual with AD Individual with neuropathologically confirmed Age at death Current age, or age of onset of AD 126 Fl 051 F1337 182 1/67 5B r§6 64 F1338 69 F1369 ^ 7 J^_7u ^ 6 130 F1550 132 F1764 176 [77 F1794 19 133 135 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0087775/manifest

Comment

Related Items