WO2011143574A2

WO2011143574A2 - Plasma biomarkers for diagnosis of alzheimer's disease

Info

Publication number: WO2011143574A2
Application number: PCT/US2011/036457
Authority: WO
Inventors: John Q. Trojanowski; Virginia M.Y. Lee; Leslie M. Shaw; Holly D. Soares; Eve H. Pickering; Andrew Kuhn; William Tzu-Lung Hu
Original assignee: The Trustees Of The University Of Pennsylvania; Pfizer Inc.
Priority date: 2010-05-14
Filing date: 2011-05-13
Publication date: 2011-11-17
Also published as: WO2011143574A3

Abstract

The present invention includes novel markers useful for diagnosing Alzheimer's disease in an individual. The markers are also useful for qualifying Alzheimer's disease status in an individual. In particular, the markers can be detected an individual's plasma sample to classify the sample as Alzheimer's disease or non- Alzheimer's disease.

Description

TITLE OF THE INVENTION

Plasma Biomarkers for Diagnosis of Alzheimer's Disease

BACKGROUND OF THE INVENTION

Alzheimer's disease is a progressive neurodegenerative disorder estimated to affect 27 million people worldwide with numbers doubling every 20 years (Prince et al, 2009 Alzheimer's Disease International 1-92). Although symptoms of Alzheimer's disease manifest early as deficits in memory and other cognitive domains, pathological data show neuropathological features of Alzheimer's disease, including amyloid plaques and neurofibrillary tangles, occur well before the onset of dementia (Sawa et al, 2009 N Engl J Med. 360(22):2302-2309). Research focused upon cognitive performance in individuals who have progressed to dementia have further specified a pre-demented stage characterized by deficits in memory and executive function (Petersen et al, 2008 CNS Spectr. 13(l):45-53). Furthermore, pre-demented stages of Alzheimer's disease appear to be accompanied by a stereotypical biomarker signature supporting the notion that Alzheimer's disease can be diagnosed prior to the onset of dementia and that clinical studies focused upon dementia prevention are feasible (Craig-Schapiro et al, 2009 Neurobiol Dis. 35(2): 128-140).

Despite recent advances in the field, the diagnosis of Alzheimer's disease has not changed dramatically over the last 25 years. Currently, two major guidelines are utilized to make a probabilistic diagnosis of dementia of the Alzheimer's type (DAT). These include criteria specified in the Diagnostic and Statistical manual of Mental Disorders, fourth edition (DSM-IV-TR) and those delineated by the National Institute of Neurological Disorders and Stroke - Alzheimer's Disease and Related Disorders (NINCDS-ADRDA). DSM-IV criteria require the presence of both a memory disorder and impairment in at least one additional cognitive domain leading to disability in social function and/or day-day activities (APA Diagnostic and Statistical Manual of Mental Disorders (TV-TR), 4th ed text revised. Washington, DC; 2000). NINCDS-ADRDA are similar, but specify that onset is gradual and other systemic or neural diseases known to impair cognition have been excluded (McKhann et al., 1984 Neurology 34(7): 939-944). Most clinical studies in Alzheimer's disease utilize either NINCDS-ADRDA or DSM-IV criteria to enroll mild to moderate stage individuals. However, it is generally recognized that individuals with dementia are already in mid to late neuropathological stages of Alzheimer's disease. It remains unclear whether Alzheimer's disease-related neuropathology is reversible once dementia manifests. Indeed, recent studies have implied that amelioration of plaque load may occur following therapeutic treatment, but such clearance may not be associated with clinical benefit once the underlying pathology is significant. These types of reports raise concern around testing disease modifying therapies in mild-moderate Alzheimer's disease where neuropathology may already be significantly advanced, and suggest that the ability to maximize clinical benefit for disease modifying strategies may rely heavily upon early intervention, prior to significant and perhaps irreversible, neuropathological sequealae.

Although the diagnosis of Alzheimer's disease, prior to the onset of dementia remains challenging, a number of recent biomarker studies have suggested that symptomatic individuals at risk of progressing to dementia do exhibit a stereotypical biomarker phenotype. In brief, symptomatic, and in some cases non-symptomatic, individuals at risk of progressing to dementia exhibit decreased levels of CSF amyloid beta peptide 42 (Αβ42) and elevated levels of tau and phosphorylated tau (Mattsson et al, 2009 Jama 302(4):385-393; Shaw et al, 2009 Ann Neurol. 65(4):403-413; Sunderland et al, 2003 Jama 289(16):2094-2103; Visser et al, 2009 Lancet Neurol. 8(7):619-627). In addition, a number of imaging studies have suggested that symptomatic individuals at risk of progressing to dementia show increased hippocampal and brain atrophy, decreased glucose metabolism and increased amyloid PET labeling (Jack et al, 2009 Brain 132(Pt 5): 1355-1365; Landau et al, 2009 Neurobiol Aging. Aug 4; Leow et al, 2009

Neuroimage). The robustness of the data have led to position papers championing the use of disease specific biomarkers, such as amyloid PET labeling or CSF Αβ42 and tau, to aid in the identification of pre-dementia Alzheimer's disease (sometimes referred to as prodromal or early Alzheimer's disease) (Dubois et al, 2007 Lancet Neurol. 6(8):734- 746). As a result, clinical trials targeted at cognitive improvement and dementia prevention have begun to enroll individuals characterized as "early Alzheimer's disease" using a combination of memory deficits and CSF or PET amyloid markers.

Unfortunately, the use of CSF and imaging biomarkers as screening tools is proving to be quite challenging especially in countries where CSF and or imaging approaches may not be standard of care. As a result, the need for an intermediate, non-invasive, cost-effective screening tool is pressing.

While CSF and imaging biomarkers hold great promise as tools to enable early Alzheimer's disease drug development, they are problematic as diagnostic tools for the general practitioner. CSF procedures in many countries are considered highly invasive and PET imaging techniques are expensive and, in some countries, not widely available. Simple cognitive tests offer an alternative screening approach, but require time that is often not reimbursed by many health care systems. In the setting of a general practitioner where time constraints with individuals are an issue, cognitive tests are often not brief enough to accommodate a standard six-ten minute individual interview. A blood test would be a simple route for screening, and, if positive, could be used in combination with more extensive cognitive testing followed by referral to a specialist for more confirmatory CSF and imaging testing that would enable early diagnosis. A simple blood test would offer a tractable option towards screening individuals at risk. Although numerous studies have attempted to identify a blood based signature for Alzheimer's disease, success has been elusive. In 2007, Ray et al. (2007 Nat Med. 13(1 1): 1359-1362), reported on a multiplex blood based panel for identifying symptomatic individuals who progress to dementia. Many of the analytes described by Ray et al. (2007 Nat Med. 13(1 1): 1359-1362), are currently available on a commercially luminex xMAP multiplex panel.

There is an unmet need for simple biochemical tests that can detect Alzheimer's disease at an early stage, monitor progression of the disease, and

discriminate between Alzheimer's disease, normal individuals, non-Alzheimer's disease dementias and other neurological disorders. The present invention fills these needs by providing among other things, novel biomarkers and combinations of biomarkers useful for diagnosing Alzheimer's disease, as well as methods and kits for using the biomarkers to diagnose Alzheimer's disease.

SUMMARY OF THE INVENTION

The present invention provides a composition for detecting Alzheimer's

Disease in an individual with at least 80% sensitivity, comprising reagents for detecting a plurality of biomarkers for Alzheimer's Disease in a biological sample from the individual. Preferably, the biological sample is a plasma sample.

In one embodiment, the biomarkers for Alzheimer's disease comprise Cortisol, pancreatic polypeptide, osteopontin, IGF BP2, and resistin.

In one embodiment, the biomarkers for Alzheimer's disease comprise resistin, pancreatic polypeptide, ApoD, G-CSF, MlPlbeta, ApoE, NrCAM, MMP2, Cortisol, PDGF-BB, MMP1, and 1-309. In one embodiment, the biomarkers for Alzheimer's disease comprise Alpha- 1 Microglobulin, Angiopoietin-2, Apolipoprotein E, Beta-2 Microglobulin, Chemokine (C-X-C motif) ligand 13 (CXCL13) also known as B lymphocyte chemoattractant (BLC), Cortisol, E-Selectin, FAS, Fatty Acid Binding Protein,

Hepatocyte Growth Factor, IGF BP-2, Interleukin 10, NT-Pro-BNP (Brain natriuretic peptide co-secreted along with a 76 amino acid N-terminal fragment), Osteopontin, Pancreatic Polypeptide, PAPP-A, Resistin, Stem Cell Factor, Tenascin C,

Thrombomodulin, tissue inhibitor of metalloproteinases-1 (TIMP-1), vascular cell adhesion molecule- 1 (VCAM-1), Vascular endothelial growth factor (VEGF), and Von Willebrand Factor.

In one embodiment, the biomarkers include one or more biomarkers selected from Table 2.

In one embodiment, the reagent for detecting a plurality of biomarkers for Alzheimer's Disease in a biological sample from the individual is an antibody. In one embodiment, the reagent is on a solid support.

The invention provides a method of diagnosing Alzheimer's Disease in an individual, the method comprising analyzing a biological sample from an individual to determine the level(s) of one or more biomarkers for Alzheimer's Disease in the sample, wherein the one or more biomarkers are selected from Table 2; and comparing the level(s) of the one or more biomarkers in the sample to Alzheimer's Disease-positive and/or Alzheimer's Disease-negative reference levels of the one or more biomarkers in order to diagnose whether the individual has Alzheimer's Disease.

In one embodiment, the method further comprises analyzing age, gender, and ApoE genotype (e2/e2, e2/e3, e2/e4, e3/e3, e3/e4 and e4/e4) of the individual in order to diagnose whether the individual has Alzheimer's Disease.

In one embodiment, the levels of the biomarkers, age, gender, and ApoE genotype are incorporated into an algorithm to diagnose whether the individual has Alzheimer's Disease.

In one embodiment, the algorithm is selected from the group consisting of Logistic Regression, Boosted Tree Models, Flexible Discriminant Analysis (FDA), K- Nearest Neighbors (KNN), Naive Bayes, Partial Least Squares (PLS), Randon Forests, Shrunken Centroids, Sparse Partial Least Squares, Support Vector Machines approaches, and any combination thereof. In one embodiment, the Alzheimer's Disease diagnosed is selected from the group consisting of late onset Alzheimer's disease, early onset Alzheimer's disease, familial Alzheimer's disease, and sporadic Alzheimer's disease.

In one embodiment, the biological sample is selected from the group consisting of whole blood, a blood component, CSF, urine, and any combination thereof.

The present invention provides a method for determining whether dementia in an individual is associated with AD, the method comprising analyzing a biological sample from an individual to determine the level(s) of Cortisol, pancreatic polypeptide, osteopontin, IGF BP2, and resistin in the sample, and comparing the level(s) of Cortisol, pancreatic polypeptide, osteopontin, IGF BP-2, and resistin in a corresponding reference sample to determine whether dementia in an individual is associated with AD.

In one embodiment, the dementia is associated with AD.

In one embodiment, the method further comprises analyzing age, gender, and ApoE genotype (e2/e2, e2/e3, e2/e4, e3/e3, e3/e4 and e4/e4) of the individual in order to determine the presence or status of dementia in the individual.

In one embodiment, the biological sample from the individual is analyzed to determine the level(s) of one or more biomarkers selected from Table 2; and comparing the level(s) of the one or more biomarkers in the sample to Alzheimer's Disease-positive and/or Alzheimer's Disease-negative reference levels of the one or more biomarkers in order to determine the presence or status of Alzheimer's Disease or other types of dementia in the individual.

In one embodiment, the levels of Cortisol, pancreatic polypeptide, osteopontin, IGF BP2, resistin biomarkers, age, gender, and ApoE genotype are incorporated into an algorithm in order to determine the presence or status of Alzheimer's Disease or other types of dementia in the individual.

The present invention also includes a method for diagnosing Alzheimer's disease in an individual, the method comprising: measuring the level of one or more biomarkers selected from Table 2 in a plasma sample from the individual, and comparing the level of the biomarkers in the plasma sample of the individual to positive control reference values and optionally to negative control reference values, wherein levels of the biomarkers in the plasma sample of the individual resemble the positive control reference values and do not resemble the negative control reference values are an indication that the individual has Alzheimer's disease. In one embodiment, the method further comprises analyzing age, gender, and ApoE genotype (e2/e2, e2/e3, e2/e4, e3/e3, e3/e4 and e4/e4) of the test individual in order to diagnose whether the individual has Alzheimer's Disease.

In one embodiment, the level of the biomarkers, age, gender, and ApoE genotype are incorporated into an algorithm to diagnose whether the individual has Alzheimer's Disease.

The present invention includes a method of determining risk of an individual developing Alzheimer's disease, the method comprising: measuring the level of one or more biomarkers selected from Table 2 in a plasma sample of an individual, and comparing the level of the biomarkers in the plasma sample of the individual to positive and optionally to negative control values, wherein levels of the biomarkers in the plasma sample of the individual that resemble the positive control and do not resemble the negative control are an indication that the individual is at risk of developing Alzheimer's disease.

In one embodiment, the method further comprises analyzing age, gender, and ApoE genotype (e2/e2, e2/e3, e2/e4, e3/e3, e3/e4 and e4/e4) of the individual in order to determine risk of an individual of developing Alzheimer's disease.

In one embodiment, the level of biomarkers, age, gender, and ApoE genotype are incorporated into an algorithm in order to determine risk of an individual of developing Alzheimer's disease.

The invention includes a method for monitoring Alzheimer's disease in an individual, the method comprising: measuring the level of one or more biomarkers selected from Table 2 in a plasma sample of an individual at a first time, and comparing the level of the biomarkers in the plasma sample of the individual at the first time to a level of the biomarkers in a plasma sample of the individual at the second time.

In one embodiment, the method further comprises analyzing age, gender, and ApoE genotype (e2/e2, e2/e3, e2/e4, e3/e3, e3/e4 and e4/e4) of the individual in order to monitor Alzheimer's disease in the individual.

In one embodiment, the level of the biomarkers, age, gender, and ApoE genotype are incorporated into an algorithm in order to monitor Alzheimer's disease in the individual.

The invention provides a method of assessing increased risk of developing Alzheimer's disease, the method comprising: obtaining a first sample of plasma from an individual blood at a first time, assessing the level of a biomarker for Alzheimer's disease in the first plasma sample to obtain a baseline level, obtaining a second sample of plasma from the individual at a second time to obtain a second level, assessing the level of the biomarker in the second plasma sample to obtain a second level, wherein if the second level is greater in the case of a biomarker that is over-expressed in AD or lower in the case of a biomarker that is under-expressed in AD, the individual is at an increased risk of developing Alzheimer's disease.

In one embodiment, the biomarker is one or more biomarkers selected from Table 2.

In one embodiment, the method further comprises analyzing age, gender, and ApoE genotype (e2/e2, e2/e3, e2/e4, e3/e3, e3/e4 and e4/e4) of the individual in order to assess whether the individual is at an increased risk of developing Alzheimer's disease.

In one embodiment, the level of the biomarkers, age, gender, and ApoE genotype are incorporated into an algorithm in order to assess whether the individual is at an increased risk of developing Alzheimer's disease.

The invention includes a method of assessing the likelihood that a pharmaceutical agent is efficacious in treating Alzheimer's disease in an individual, the method comprising: obtaining a first sample of plasma from an individual in the absence of a pharmaceutical agent, assessing the level of a biomarker for Alzheimer's disease in the first plasma sample to obtain a baseline level, administering the pharmaceutical agent to the individual, obtaining a second sample of plasma from the individual after administration of the pharmaceutical agent, assessing the level of the biomarker for Alzheimer's disease in the second plasma sample to obtain a treated level, wherein the likelihood that the pharmaceutical agent treats Alzheimer's disease is increased if the treated level is lower than the baseline level in the case of a biomarker that is over- expressed in AD and higher than the baseline level in the case of a biomarker that is under-expressed in AD.

In one embodiment, the method further comprises analyzing age, gender, and ApoE genotype (e2/e2, e2/e3, e2/e4, e3/e3, e3/e4 and e4/e4) of the individual in order to assess the likelihood that the pharmaceutical agent is efficacious in treating

Alzheimer's disease in the individual. In one embodiment, the levels of the biomarkers, age, gender, and ApoE genotype are incorporated into an algorithm in order to assess the likelihood that the pharmaceutical agent is efficacious in treating Alzheimer's disease in the individual.

The invention includes a method of diagnosing Alzheimer's Disease in an individual, comprising obtaining a plasma sample from the individual; assaying levels of at least five biomarkers in the plasma sample, wherein the at least five biomarkers comprise: Cortisol, IGF BP2, pancreatic polypeptide, resistin, and ApoE; and comparing the levels of each of the at least five biomarkers to non-AD controls, wherein a statistically significant increase in the levels of Cortisol, IGF BP2, pancreatic polypeptide, and resistin, and a statistically significant decrease in the level of ApoE provides a positive diagnosis of AD in the individual.

The invention includes a method of diagnosing Alzheimer's Disease in an individual, comprising obtaining a plasma sample from the individual; assaying levels of at least five biomarkers in the plasma sample, wherein the at least five biomarkers comprise: Cortisol, IGF BP2, pancreatic polypeptide, resistin, and osteopontin; and comparing the levels of each of the at least five biomarkers to non-AD controls, wherein a statistically significant increase in the levels of Cortisol, IGF BP2, pancreatic polypeptide, osteopontin and resistin provides a positive diagnosis of AD in the individual.

In one embodiment, the method further comprises conducting one or more cognitive tests on the individual to confirm the positive diagnosis of AD.

In one embodiment, the method further comprises obtaining a CSF sample from the individual to confirm the positive diagnosis of AD.

The invention provides a method for screening to identify individuals at increased risk of developing AD, the method comprising obtaining a plasma sample from the individual; assaying levels of at least five biomarkers in the plasma sample, wherein the at least five biomarkers comprise: Cortisol, IGF BP2, pancreatic polypeptide, resistin, and ApoE; and comparing the levels of each of the at least five biomarkers to non-AD controls, wherein a statistically significant increase in the levels of Cortisol, IGF BP2, pancreatic polypeptide, and resistin, and a statistically significant decrease in the level of ApoE provides an identification of an individual at increased risk of developing AD.

The invention provides a method for screening to identify individuals at increased risk of developing AD, the method comprising obtaining a plasma sample from the individual; assaying levels of at least five biomarkers in the plasma sample, wherein the at least five biomarkers comprise: Cortisol, IGF BP2, pancreatic polypeptide, resistin, and osteopontin; and comparing the levels of each of the at least five biomarkers to non- AD controls, wherein a statistically significant increase in the levels of Cortisol, IGF BP2, pancreatic polypeptide, osteopontin and resistin provides an identification of an individual at increased risk of developing AD.

The invention provides a method of diagnosing Alzheimer's Disease in an individual, comprising obtaining a plasma sample from the individual; assaying levels of at least five biomarkers in the plasma sample, wherein the at least five biomarkers comprise: Cortisol, IGF BP2, pancreatic polypeptide, resistin, and ApoE; and determining whether, relative to non-AD controls, the levels of the at least five biomarkers provide a signature of a positive diagnosis of AD in the individual, wherein the signature comprises: a statistically significant increase in the levels of each of Cortisol, IGF BP2, pancreatic polypeptide, and resistin, and a statistically significant decrease in the level of ApoE.

The invention provides a method of diagnosing Alzheimer's Disease in an individual, comprising obtaining a plasma sample from the individual; assaying levels of at least five biomarkers in the plasma sample, wherein the at least five biomarkers comprise: Cortisol, IGF BP2, pancreatic polypeptide, resistin, and osteopontin; and determining whether, relative to non-AD controls, the levels of the at least five biomarkers provide a signature of a positive diagnosis of AD in the individual, wherein the signature comprises: a statistically significant increase in the levels of each of Cortisol, IGF BP2, pancreatic polypeptide, osteopontin and resistin.

The invention provides a kit for detecting Alzheimer's Disease in an individual with at least 80% sensitivity, comprising a reagent for detecting a plurality of biomarkers for Alzheimer's Disease in a biological sample from the individual.

In one embodiment, the reagent is capable of detecting biomarkers of

Alzheimer's disease in a plasma sample.

In one embodiment, the biomarkers of Alzheimer's disease comprise Cortisol, pancreatic polypeptide, osteopontin, IGF BP2, and resistin.

In one embodiment, the biomarkers for Alzheimer's disease comprise resistin, pancreatic polypeptide, ApoD, G-CSF, MlPlbeta, ApoE, NrCAM, MMP2, Cortisol, PDGF-BB, MMP1, and 1-309.

In one embodiment, the biomarkers for Alzheimer's disease comprise Alpha- 1 Microglobulin, Angiopoietin-2, Apolipoprotein E, Beta-2 Microglobulin, Chemokine (C-X-C motif) ligand 13 (CXCL13) also known as B lymphocyte chemoattractant (BLC), Cortisol, E-Selectin, FAS, Fatty Acid Binding Protein,

In one embodiment, the reagent is an antibody.

In one embodiment, the reagent is on a solid support.

In one embodiment, the kit further comprises an instruction manual.

BRIEF DESCRIPTION OF THE DRAWINGS

For the purpose of illustrating the invention, there are depicted in the drawings certain embodiments of the invention. However, the invention is not limited to the precise arrangements and instrumentalities of the embodiments depicted in the drawings.

Figure 1 is an image depicting Test set ROC curves for each model. The solid black dot indicates the default cutoff of 0.50 and the square indicates the alternate cutoff to get sensitivity closest to 0.80. ROC curves used 190 analytes and age, gender and ApoE4 allele and top 30 analytes from Partial Least Square Model (PLS).

Figure 2 is a series of graphs depicting expression of analytes in CSF Autopsy Confirmed Samples. Graphs represent means and error bars represent SEM.

Figure 3 is a graph depicting ROC curves using logistic regression of baseline (age, gender and ApoE genotype) vs Rules-Based Medicine (RBM) analytes alone versus baseline plus 24 RBM analytes. Thirty two total features in the final model: Age, gender, ApoE genotype (e2/e2, e2/e3, e2/e4, e3/e3, e3/e4 and e4/e4), alphal microglobulin, Apolipoprotein E, angiopoietin-2, beta2 macroglobulin, BLC, Cortisol, E- selectin, Fatty acid binding protein, FAS, HGF, IGFBP-2, IL-10, NT proBNP,

Osteopontin, Pancreatic polypeptide, PAPP-A, Resistin, Stem Cell Factor, Tenascin C, Thrombomodulin, TIMP-1, VCAM, VEGF and von Willebrand Factor.

Figure 4A is an image demonstrating a correlation matrix analysis in the Alzheimer's disease samples exhibiting a high degree of correlation amongst the tested analytes. Figure 4B is a chart demonstrating high correlation of the analytes with each other and that there is high correlation with other analytes on the panel. For example, NT-proBNP showed high correlation with IGFBP2 and beta2 microglobulin suggesting these analytes may be linked in an AD specific pathophysiological pathway. Without wishing to be bound by any particular theory, it is believed that analytes that exhibit high correlation can be interchanged with other highly correlated analytes and still obtain similar differentiation performance.

Figure 5 is an image demonstrating that plasma levels of ApoE, CRP and IL-15 differ dependent upon genetic ApoE allele status independent of diagnosis. Levels of ApoE4 are lowest in subjects with one or more E4 allele(s) and highest in subjects with one or more E2 allele(s).

Figure 6 is an image depicting graphical summary of the training and test set estimates for sensitivity and specificity

Figure 7 is an image depicting test set sensitivity and specificity profiles across various cutoffs for each model.

Figure 8 is an image depicting the overall distribution of PLS variables importance scores.

Figure 9 is an image depicting the correlation matrix for the top 15 predictors.

Figure 10 is an image depicting the training set data points for the top 15 predictors.

DETAILED DESCRIPTION OF THE INVENTION

The present invention provides a method for diagnosing, monitoring and/or staging neurological disorders comprising the use of a multivariate and/or univariate approach to identify a neurological disorder in an individual.

The present invention relates generally to diagnostic methods and markers, prognostic methods and markers, and therapy evaluators for neurodegenerative disorders, such as Alzheimer's Disease. In one embodiment, the markers of the invention are useful for detecting early stage Alzheimer's disease.

In certain embodiments, the method comprises the step of obtaining a sample of plasma from the individual's blood, and assessing the level of markers of Alzheimer's Disease in the plasma sample. Thus, the present invention relates to markers of Alzheimer's Disease, methods for diagnosis of Alzheimer's Disease, methods of determining predisposition to Alzheimer's Disease, methods of monitoring

progression/regression of Alzheimer's Disease, methods of assessing efficacy of compositions for treating Alzheimer's Disease, methods of screening compositions for activity in modulating markers of Alzheimer's Disease, methods of treating Alzheimer's Disease, as well as other methods based on markers of Alzheimer's Disease.

In certain embodiments, the invention further provides methods for permitting refinement of disease diagnosis, disease risk prediction, and clinical management of individuals associated with a neurodegenerative disorder. The

Alzheimer's disease markers of the invention represent a plasma based panel for assessing Alzheimer's disease that can be used for determining the disease state or disease risk. The detection of the selective markers of the invention in individuals, or samples obtained therefrom permits refinement of disease diagnosis, disease risk prediction, and clinical management of individuals being treated with agents that are associated with Alzheimer's disease.

In one embodiment, the Alzheimer's disease markers of the invention include one or more of the markers shown in Table 2.

In one embodiment of the invention, age, gender, and ApoE genotype (e2/e2, e2/e3, e2/e4, e3/e3, e3/e4 and e4/e4) are additional factors that are considered in identifying an individual for Alzheimer's disease.

In one embodiment of the invention, markers of AD includes at least

Cortisol, pancreatic polypeptide, osteopontin, IGF BP2, and resistin.

In one embodiment of the invention, markers of AD includes at least resistin, e3/e3, pancreatic polypeptide, e3/e4, ApoD, G-CSF, MlPlbeta, ApoE, NrCAM, MMP2, Cortisol, PDGF-BB, e2/e3, MMP 1, and 1-309.

In one embodiment of the invention, markers of AD includes at least

Alpha- 1 Microglobulin, Angiopoietin-2, Apolipoprotein E, Beta-2 Microglobulin, Chemokine (C-X-C motif) ligand 13 (CXCL13) also known as B lymphocyte chemoattractant (BLC), Cortisol, E-Selectin, FAS, Fatty Acid Binding Protein,

Thrombomodulin, tissue inhibitor of metalloproteinases-1 (TIMP-1), vascular cell adhesion molecule- 1 (VCAM-1), Vascular endothelial growth factor (VEGF), and Von Willebrand Factor. In one embodiment, the plasma based panel for assessing Alzheimer's disease comprises one or more of the following markers: Alpha-1 Microglobulin, Angiopoietin-2, Apolipoprotein E, Beta-2 Microglobulin, Chemokine (C-X-C motif) ligand 13 (CXCL13) also known as B lymphocyte chemoattractant (BLC), Cortisol, E- Selectin, FAS, Fatty Acid Binding Protein, Hepatocyte Growth Factor, IGF BP-2,

Interleukin 10, NT-Pro-BNP (Brain natriuretic peptide co-secreted along with a 76 amino acid N-terminal fragment), Osteopontin, Pancreatic Polypeptide, PAPP-A, Resistin, Stem Cell Factor, Tenascin C, Thrombomodulin, tissue inhibitor of metalloproteinases- 1 (TIMP-1), vascular cell adhesion molecule- 1 (VCAM-1), Vascular endothelial growth factor (VEGF), and Von Willebrand Factor. In some embodiments, age, gender, ApoE genotype (e2/e2, e2/e3, e2/e4, e3/e3, e3/e4 and e4/e4) are additional factors that are considered in evaluating the diagnosis of an individual for Alzheimer's disease.

The invention includes a method comprising obtaining a sample from an individual, and assessing the level of one or more of Alpha-1 Microglobulin,

Angiopoietin-2, Apolipoprotein E, Beta-2 Microglobulin, BLC, Cortisol, E-Selectin, FAS, Fatty Acid Binding Protein, Hepatocyte Growth Factor, IGF BP-2, Interleukin 10, NT-Pro-BNP, Osteopontin, Pancreatic Polypeptide, PAPP-A, Resistin, Stem Cell Factor, Tenascin C, Thrombomodulin, TIMP-1, VCAM-1, VEGF, and Von Willebrand Factor in the sample. However, the invention should not be limited to only these markers disclosed herein (e.g., Table 2) because a skilled artisan when armed with the present disclosure would be able identify additional markers that can be used as indicators for Alzheimer's disease. For example, as discussed in the Examples, a test sample and a control sample can be subjected to any commercially available panel comprising a plurality of markers and analyzed according to the statistic models disclosed herein to identify markers associated with AD.

Moreover, the disclosure presented here demonstrates a high degree of correlation amongst certain markers for identifying AD in an individual. A skilled artisan when armed with the present disclosure, particularly Figures 4A and 4B, would recognize that analytes that exhibit high correlation can be interchanged with other highly correlated analytes and still obtain similar differentiation performance. For example, NT-proBNP showed high correlation with IGFBP2 and beta2 microglobulin suggesting these analytes may be linked in an AD specific pathophysiological pathway. Thus, markers that are associated with an AD specific pathophysiological pathway can be interchangeable. Accordingly, correlation amongst the markers of the invention provides means to identify other related markers associated with the specific pathophysiological pathway as being indicators for Alzheimer's disease.

Any number of biomarkers may be used in the methods disclosed herein. That is, the disclosed methods may include the determination of the level(s) of one biomarker, two or more biomarkers, three or more biomarkers, four or more biomarkers, five or more biomarkers, six or more biomarkers, seven or more biomarkers, eight or more biomarkers, nine or more biomarkers, ten or more biomarkers, fifteen or more biomarkers, etc., including a combination of all of the biomarkers in each or all of Tables 2, 3, 7, 8, 9, 10, 1 1, 12 or any fraction thereof. In another aspect, the biomarkers are used in combination with other factors that predict for AD. Such factors include but are not limited to age, gender, ApoE genotype (e2/e2, e2/e3, e2/e4, e3/e3, e3/e4 and e4/e4).

In one embodiment, an immunoassay is used for the assessment of a marker level. In another embodiment, a luminex technology multiplex immunoassay is used to assess the marker level.

In another embodiment, a method of diagnosing Alzheimer's disease in an individual comprises the steps of obtaining a first sample of plasma from the individual at a first time; assessing the level of Alzheimer's disease markers using plasma based panel of the invention in the first plasma sample to obtain a baseline level; obtaining a second sample of plasma from the individual at a second time and assessing the level of

Alzheimer's disease markers in the second plasma sample to obtain a second level. If the second level is significantly altered compared to the baseline level, the individual is at an increased risk of developing or having Alzheimer's disease. In one embodiment, the second level is also compared to a reference population of an individual's without Alzheimer's disease; if the second level is significantly altered compared to the level derived from a reference population, the individual is at an increased risk of developing or having Alzheimer's disease.

In still further embodiments, the invention provides methods of monitoring a plasma panel of particular markers of Alzheimer's disease to evaluate the progress of a therapeutic treatment of Alzheimer's disease.

In another embodiment, the invention provides methods for selecting a patient that is most likely to respond to treatment.

The invention also provides methods for screening an individual to determine if the individual is at increased risk of having Alzheimer's disease. Individuals found to be at increased risk can be given appropriate therapy and monitored using the methods of the invention.

Other methods and kits useful in practicing the methods of the invention are provided herein.

Definitions

As used herein, each of the following terms has the meaning associated with it in this section.

The articles "a" and "an" are used herein to refer to one or to more than one (i.e., to at least one) of the grammatical object of the article. By way of example, "an element" means one element or more than one element.

"About" as used herein when referring to a measurable value such as an amount, a temporal duration, and the like, is meant to encompass variations of ±20% or ±10%, more preferably ±5%, even more preferably ±1%, and still more preferably ±0.1% from the specified value, as such variations are appropriate to perform the disclosed methods.

The term "abnormal" when used in the context of organisms, tissues, cells or components thereof, refers to those organisms, tissues, cells or components thereof that differ in at least one observable or detectable characteristic (e.g., age, treatment, time of day, etc.) from those organisms, tissues, cells or components thereof that display the "normal" (expected) respective characteristic. Characteristics which are normal or expected for one cell or tissue type, might be abnormal for a different cell or tissue type.

The term "antibody," as used herein, refers to an immunoglobulin molecule which is able to specifically bind to a specific epitope on an antigen.

Antibodies can be intact immunoglobulins derived from natural sources or from recombinant sources and can be immunoreactive portions of intact immunoglobulins. The antibodies in the present invention may exist in a variety of forms including, for example, polyclonal antibodies, monoclonal antibodies, intracellular antibodies

("intrabodies"), Fv, Fab and F(ab)2, as well as single chain antibodies (scFv), heavy chain antibodies, such as camelid antibodies, and humanized antibodies (Harlow et al, 1999, Using Antibodies: A Laboratory Manual, Cold Spring Harbor Laboratory Press, NY; Harlow et al, 1989, Antibodies: A Laboratory Manual, Cold Spring Harbor, New York; Houston et al, 1988, Proc. Natl. Acad. Sci. USA 85:5879-5883; Bird et al, 1988, Science 242:423-426). By the term "synthetic antibody" as used herein, is meant an antibody which is generated using recombinant DNA technology, such as, for example, an antibody expressed by a bacteriophage as described herein. The term should also be construed to mean an antibody which has been generated by the synthesis of a DNA molecule encoding the antibody and which DNA molecule expresses an antibody protein, or an amino acid sequence specifying the antibody, wherein the DNA or amino acid sequence has been obtained using synthetic DNA or amino acid sequence technology which is available and well known in the art.

As used herein, the term "heavy chain antibody" or "heavy chain antibodies" comprises immunoglobulin molecules derived from camelid species, either by immunization with a peptide and subsequent isolation of sera, or by the cloning and expression of nucleic acid sequences encoding such antibodies. The term "heavy chain antibody" or "heavy chain antibodies" further encompasses immunoglobulin molecules isolated from an animal with heavy chain disease, or prepared by the cloning and expression of VH (variable heavy chain immunoglobulin) genes from an animal.

As used herein, an "immunoassay" refers to any binding assay that uses an antibody capable of binding specifically to a target molecule to detect and quantify the target molecule.

By the term "specifically binds," as used herein with respect to an antibody, is meant an antibody which recognizes an specific antigen, but does not substantially recognize or bind other molecules in a sample. For example, an antibody that specifically binds to an antigen from one species may also bind to that antigen from one or more species. But, such cross-species reactivity does not itself alter the classification of an antibody as specific. In another example, an antibody that specifically binds to an antigen may also bind to different allelic forms of the antigen. However, such cross reactivity does not itself alter the classification of an antibody as specific.

In some instances, the terms "specific binding" or "specifically binding", can be used in reference to the interaction of an antibody, a protein, or a peptide with a second chemical species, to mean that the interaction is dependent upon the presence of a particular structure (e.g., an antigenic determinant or epitope) on the chemical species; for example, an antibody recognizes and binds to a specific protein structure rather than to proteins generally. If an antibody is specific for epitope "A", the presence of a molecule containing epitope A (or free, unlabeled A), in a reaction containing labeled "A" and the antibody, will reduce the amount of labeled A bound to the antibody. As used herein, "biomarker" in the context of the present invention encompasses, without limitation, proteins, nucleic acids, and metabolites, together with their polymorphisms, mutations, variants, modifications, subunits, fragments, protein- ligand complexes, and degradation products, protein-ligand complexes, elements, related metabolites, and other analytes or sample-derived measures. Biomarkers can also include mutated proteins or mutated nucleic acids. Biomarkers also encompass non-blood borne factors or non-analyte physiological markers of health status, such as clinical parameters, as well as traditional laboratory risk factors. Biomarkers also include any calculated indices created mathematically or combinations of any one or more of the foregoing measurements, including temporal trends and differences.

As used herein, the term "data" in relation to one or more biomarkers, or the term "biomarker data" generally refers to data reflective of the absolute and/or relative abundance (level) of a product of a biomarker in a sample. As used herein, the term "dataset" in relation to one or more biomarkers refers to a set of data representing levels of each of one or more biomarker products of a panel of biomarkers in a reference population of subjects. A dataset can be used to generate a formula/classifier of the invention. According to one embodiment the dataset need not comprise data for each biomarker product of the panel for each individual of the reference population. For example, the "dataset" when used in the context of a dataset to be applied to a formula can refer to data representing levels of products of each biomarker for each individual in one or more reference populations, but as would be understood can also refer to data representing levels of products of each biomarker for 99%, 95%, 90%, 85%, 80%, 75%, 70% or less of the individuals in each of said one or more reference populations and can still be useful for purposes of applying to a formula.

"Differentially increased expression" or "up regulation" refers to biomarker product levels which are at least 10% or more, for example, 20%, 30%, 40%, or 50%, 60%, 70%, 80%, 90% higher or more, and/or 1.1 fold, 1.2 fold, 1.4 fold, 1.6 fold, 1.8 fold higher or more, than a control.

"Differentially decreased expression" or "down regulation" refers to biomarker product levels which are at least 10% or more, for example, 20%, 30%, 40%, or 50%, 60%, 70%, 80%, 90% lower or less, and/or 0.9 fold, 0.8 fold, 0.6 fold, 0.4 fold, 0.2 fold, 0.1 fold or less lower than a control.

A "formula," "algorithm," or "model" is any mathematical equation, algorithmic, analytical or programmed process, or statistical technique that takes one or more continuous or categorical inputs (herein called "parameters") and calculates an output value, sometimes referred to as an "index" or "index value." Non-limiting examples of "formulas" include sums, ratios, and regression operators, such as coefficients or exponents, biomarker value transformations and normalizations (including, without limitation, those normalization schemes based on clinical parameters, such as gender, age, or ethnicity), rules and guidelines, statistical classification models, and neural networks trained on historical populations. Of particular use in combining Alzheimer's disease markers and other biomarkers are linear and non-linear equations and statistical classification analyses to determine the relationship between levels of Alzheimer's disease markers detected in a subject sample. In panel and combination construction, of particular interest are structural and synactic statistical classification algorithms, and methods of risk index construction, utilizing pattern recognition features, including established techniques such as cross-correlation, Principal Components Analysis (PCA), factor rotation, Logistic Regression (LogReg), Linear Discriminant Analysis (LDA), Linear Discriminant Analysis (LDA), Support Vector Machines (SVM), Random Forest (RF), Partial Least Squares, Sparse Partial Least Squares, Flexible Discriminant Analysis, Recursive Partitioning Tree (RPART), as well as other related decision tree classification techniques, Nearest Shrunken Centroids (SC)", stepwise model selection procedures, Kth-Nearest Neighbor, Boosting or Boosted Tree, Decision Trees, Neural Networks, Bayesian Networks, Support Vector Machines, and Hidden Markov Models, among others. Other techniques may be used in survival and time to event hazard analysis, including Cox, Weibull, Kaplan-Meier and Greenwood models well known to those of skill in the art.

An "analyte", as used herein refers to any substance or chemical constituent that is undergoing analysis. For example, an "analyte" can refer to any atom and/or molecule; including their complexes and fragment ions. The term may refer to a single component or a set of components. In the case of biological

molecules/macromolecules, such analytes include but are not limited to: polypeptides, polynucleotides, proteins, peptides, antibodies, DNA, RNA, carbohydrates, steroids, and lipids, and any detectable moiety thereof, e.g. immunologically detectable fragments.

A "disease" is a state of health of an animal wherein the animal cannot maintain homeostasis, and wherein if the disease is not ameliorated then the animal's health continues to deteriorate. A "disorder" in an animal is a state of health in which the animal is able to maintain homeostasis, but in which the animal's state of health is less favorable than it would be in the absence of the disorder. Left untreated, a disorder does not necessarily cause a further decrease in the animal's state of health.

As used herein, the terms "Alzheimer's disease" and "Alzheimer's disease" refer to a neurodegenerative disorder and encompass familial Alzheimer's disease and sporadic Alzheimer's disease. The term "familial Alzheimer's disease" refers to

Alzheimer's disease associated with genetic factors (i.e., inheritance is demonstrated) while "sporadic Alzheimer's disease" refers to Alzheimer's disease that is not associated with prior family history of the disease. Symptoms indicative of Alzheimer's disease in human subjects typically include, but are not limited to, mild to severe dementia, progressive impairment of memory (ranging from mild forgetfulness to disorientation and severe memory loss), poor visual spatial skills, personality, changes, poor impulse control, poor judgment, distrust of others, increased stubbornness, restlessness, poor planning ability, poor decision making, and social withdrawal. In severe cases, patients lose the ability to use language and communicate, and require assistance in personal hygiene, eating and dressing, and are eventually bedridden. Hallmark pathologies within brain tissue include extracellular neuritic amyloid plaques, neurofibrillary tangles, neurofibrillary degeneration, granulovascular neuronal degeneration, synaptic loss, and extensive neuronal cell death.

"Increased risk of developing Alzheimer's disease" is used herein to refer to an increase in the likelihood or possibility of developing Alzheimer's disease. This risk can be assessed relative to an individual's own risk, or with respect to a reference population that does not have clinical evidence of Alzheimer's disease. The reference population may be representative of the individual with regard to approximate age, age group and/or gender.

"Delaying development of Alzheimer's disease" as used herein refers to a prolonging of the time to the development of Alzheimer's disease and/or delay in the rate of increased extent of Alzheimer's disease.

"Alleviating Alzheimer's disease" as used herein refers to a decrease in the severity of Alzheimer's disease.

As used herein, the terms "Alzheimer's patient", "Alzheimer's disease patient", and "individual diagnosed with Alzheimer's disease" all refer to an individual who has been diagnosed with Alzheimer's disease or has been given a probable diagnosis of Alzheimer's Disease.

An "individual with mild Alzheimer's disease" is an individual who has been diagnosed with Alzheimer's disease or has been given a diagnosis of probable Alzheimer's disease. In some instances, an "individual with mild Alzheimer's disease" has either been assessed with the Mini-Mental State Examination (MMSE) (referenced in Folstein et al, J. Psychiatr. Res 1975; 12: 1289-198) and scored 22-27 or would achieve a score of 22-27 upon MMSE testing.

An "individual with moderate Alzheimer's disease" is an individual who has been diagnosed with Alzheimer's disease or has been given a diagnosis of probable Alzheimer's disease. In some instances, an "individual with moderate Alzheimer's disease" has either been assessed with the MMSE and scored 16-21 or would achieve a score of 16-21 upon MMSE testing.

An "individual with severe Alzheimer's disease" is an individual who has been diagnosed with Alzheimer's disease or has been given a diagnosis of probable

Alzheimer's disease. In some instances, an "individual with severe Alzheimer's disease" has either been assessed with the MMSE and scored 12-15 or would achieve a score of 12-15 upon MMSE testing.

The "level" of one or more biomarkers means the absolute or relative amount or concentration of the biomarker in the sample.

"Measuring" or "measurement," or alternatively "detecting" or

"detection," means assessing the presence, absence, quantity or amount (which can be an effective amount) of either a given substance within a clinical or subject-derived sample, including the derivation of qualitative or quantitative concentration levels of such substances, or otherwise evaluating the values or categorization of a subject's clinical parameters.

A "reference level" of a biomarker means a level of the biomarker that is indicative of a particular disease state, phenotype, or lack thereof, as well as combinations of disease states, phenotypes, or lack thereof. A "positive" reference level of a biomarker means a level that is indicative of a particular disease state or phenotype. A "negative" reference level of a biomarker means a level that is indicative of a lack of a particular disease state or phenotype. For example, an "Alzheimer's Disease-positive reference level" of a biomarker means a level of a biomarker that is indicative of a positive diagnosis of Alzheimer's Disease in a subject, and an "Alzheimer's Disease-negative reference level" of a biomarker means a level of a biomarker that is indicative of a negative diagnosis of Alzheimer's Disease in a subject. As another example, an

"Alzheimer's-Disease-progression-positive reference level" of a biomarker means a level of a biomarker that is indicative of progression of Alzheimer's Disease in a subject, and an "Alzheimer's-Disease-regression-positive reference level" of a biomarker means a level of a biomarker that is indicative of regression of the Alzheimer's Disease. A "reference level" of a biomarker may be an absolute or relative amount or concentration of the biomarker, a presence or absence of the biomarker, a range of amount or concentration of the biomarker, a minimum and/or maximum amount or concentration of the biomarker, a mean amount or concentration of the biomarker, and/or a median amount or concentration of the biomarker; and, in addition, "reference levels" of combinations of biomarkers may also be ratios of absolute or relative amounts or concentrations of two or more biomarkers with respect to each other. Appropriate positive and negative reference levels of biomarkers for a particular disease state, phenotype, or lack thereof may be determined by measuring levels of desired biomarkers in one or more appropriate subjects, and such reference levels may be tailored to specific populations of subjects (e.g., a reference level may be age-matched so that comparisons may be made between biomarker levels in samples from subjects of a certain age and reference levels for a particular disease state, phenotype, or lack thereof in a certain age group). Such reference levels may also be tailored to specific techniques that are used to measure levels of biomarkers in biological samples, where the levels of biomarkers may differ based on the specific technique that is used. Such reference levels may also be tailored to specific statistical models used to analyze levels of biomarkers in biological samples, where the levels of biomarkers may differ based on the specific statistical model that is used.

As used herein, a "detector molecule" is a molecule that may be used to detect a compound of interest. Non-limiting examples of a detector molecule are molecules that bind specifically to a compound of interest, such as, but not limited to, an antibody, a cognate receptor, and a small molecule.

An "effective amount" or "therapeutically effective amount" of a compound is that amount of compound which is sufficient to provide a beneficial effect to the individual to which the compound is administered. An "effective amount" of a delivery vehicle is that amount sufficient to effectively bind or deliver a compound.

As used herein, an "instructional material" includes a publication, a recording, a diagram, or any other medium of expression which can be used to communicate the usefulness of a compound, composition, vector, or delivery system of the invention in the kit for effecting alleviation of the various diseases or disorders recited herein. Optionally, or alternately, the instructional material can describe one or more methods of alleviating the diseases or disorders in a cell or a tissue of a individual. The instructional material of the kit of the invention can, for example, be affixed to a container which contains the identified compound, composition, vector, or delivery system of the invention or be shipped together with a container which contains the identified compound, composition, vector, or delivery system. Alternatively, the instructional material can be shipped separately from the container with the intention that the instructional material and the compound be used cooperatively by the recipient.

The term "microarray" refers broadly to both "DNA microarrays" and "DNA chip(s)," and encompasses all art-recognized solid supports, and all art-recognized methods for affixing nucleic acid molecules thereto or for synthesis of nucleic acids thereon.

The terms "patient," "subject," "individual," and the like are used interchangeably herein, and refer to any animal, or cells thereof whether in vitro or in situ, amenable to the methods described herein. In certain non-limiting embodiments, the patient, subject or individual is a human.

The term "protein" typically refers to large polypeptides.

"Sample" or "biological sample" as used herein means a biological material isolated from an individual. The biological sample may contain any biological material suitable for detecting the desired biomarkers, and may comprise cellular and/or non-cellular material obtained from the individual.

A "therapeutic" treatment is a treatment administered to an individual who exhibits signs of pathology, for the purpose of diminishing or eliminating those signs.

The term to "treat," as used herein, means reducing the frequency with which symptoms are experienced by an individual or subject or administering an agent or compound to reduce the frequency with which symptoms are experienced. In some instances, "treat" and "treating" are not limited to the case where the subject (e.g. patient) is cured and the disease is eradicated. Rather, the present invention also contemplates treatment that merely reduces symptoms, improves (to some degree) and/or delays disease progression. The term "treatment" also refers to the alleviation, amelioration, and/or stabilization of symptoms, as well as delay in progression of symptoms of a particular disorder. For example, "treatment" of Alzheimer's disease includes any one or more of: elimination of one or more symptoms of Alzheimer's disease, reduction of one or more symptoms of Alzheimer's disease, stabilization of the symptoms of Alzheimer's disease (e.g., failure to progress to more advanced stages of Alzheimer's disease), and delay in progression (i.e., worsening) of one or more symptoms of Alzheimer's disease.

As used herein, "treating a disease or disorder" means reducing the frequency with which a symptom of the disease or disorder is experienced by an individual. Disease and disorder are used interchangeably herein.

Ranges: throughout this disclosure, various aspects of the invention can be presented in a range format. It should be understood that the description in range format is merely for convenience and brevity and should not be construed as an inflexible limitation on the scope of the invention. Accordingly, the description of a range should be considered to have specifically disclosed all the possible subranges as well as individual numerical values within that range and, when appropriate, partial integers of the numerical values within ranges. For example, description of a range such as from 1 to 6 should be considered to have specifically disclosed subranges such as from 1 to 3, from 1 to 4, from 1 to 5, from 2 to 4, from 2 to 6, from 3 to 6 etc., as well as individual numbers within that range, for example, 1, 2, 2.7, 3, 4, 5, 5.3, and 6. This applies regardless of the breadth of the range. Description

Alzheimer's disease is a progressive neurodegenerative disorder characterized by memory and cognitive impairments and other non-cognitive behavioral symptoms. Age is the strongest risk factor, wherein almost 50% of people over the age of 85 are affected. Early-onset Alzheimer's disease (EOAD) is associated with genetic mutations in amyloid precursor protein (APP), presenilin 1 (PSE 1) and presenilin 2

(PSEN2). However, sporadic or late-onset Alzheimer's disease (LOAD) is multi-factorial and genetically more complex. In addition, genetic factors may account for as much as 80% of the disease risk associated with LOAD (Gatz et al. (2006) Arch. Gen. Psychiatry 63(2): 168-174). While monogenic mutations cause EOAD, the only extensively validated susceptibility gene for LOAD is the apolipoprotein E (ΑΡΟΕ-ε4) allele

(Saunders et al. (1993) Neurology 43(8): 1467-1472 and Farrer et al. (1997) JAMA 278(16): 1349-1356). But alleles of the APOE gene do not account for all of the genetic load responsible for LOAD predisposition. Currently, Alzheimer's can be identified through the use of cognitive tests combined with cerebrospinal fluid (CSF) and imaging biomarkers. However, such tests are often costly and invasive and are difficult to implement as screening tools to identify patient s suitable for clinical trial study.

The present invention provides novel Alzheimer's disease biomarkers present in a biological sample of an individual. Preferably, the biological sample is plasma. The level of these biomarkers allow a more accurate diagnosis or prognosis of Alzheimer's disease in individuals that are at risk for Alzheimer's disease, that exhibit no clinical signs of Alzheimer's disease, or that exhibit minor clinical signs of Alzheimer's disease. Furthermore, the measurement of the biomarkers of the invention allows the monitoring of Alzheimer's disease, such that a comparison of biomarker levels facilitates an evaluation of disease progression in individuals that have been diagnosed with Alzheimer's disease, or that do not yet exhibit any clinical signs of Alzheimer's disease. Moreover, the Alzheimer's disease biomarkers of the invention may be used in concert with known Alzheimer's disease biomarkers such that a more accurate diagnosis or prognosis of Alzheimer's disease may be made.

In one embodiment, the invention provides a screening tool to detect Alzheimer's disease in an individual. Preferably, the screening tool includes a examining a multiplexed plasma panel to differentiate Alzheimer's disease from age-matched controls. In some instances, the screening tool encompasses a multivariate based approach combined a statistical model to predict Alzheimer's disease in an individual. For example, the multivariate approach useful in the invention encompasses the use of a multiplexed plasma panel in combination with age, gender and ApoE genetic status (6 co- variates e2/e2; e2/e3; e2/e4; e3/e4; e3/e3; e4/e4) for determining Alzheimer's disease in an individual. In some instances, the screening tool of the invention can differentiate Alzheimer's disease from age-matched controls with high sensitivity suitable for preliminary screening of individuals who might be eligible for clinical trial enrollment in early Alzheimer's disease studies. The disclosure presented herein also includes additional markers associated with increased or decreased risk of Alzheimer's disease as shown in Table 2.

The present invention relates partly on the discovery of biomarkers that can differentiate Alzheimer's disease from controls and identify individuals at risk of progressing to dementia. A two tiered strategy was employed to find the optimal discriminatory panel of biomarkers. First, multivariate approaches were utilized to identify top plasma analytes that contributed to maximal differentiation using a training/test set approach. Second, plasma analytes were compared to expression in CSF from individuals with diagnosis confirmed by autopsy analysis. A top list of analytes was identified based upon 1) performance in multivariate models, 2) expression in CSF from autopsy confirmed individuals and 3) upon linkage to disease by pathway analysis. A final analyte panel of analytes, combined with age, gender and ApoE allele status, was then subjected to logistic regression analysis to determine performance in differentiating Alzheimer's disease from controls. An objective was to determine whether a plasma based signature with a sensitivity of at least 80% could be achieved. In some instances, the objective was to determine whether a plasma based signature with a sensitivity of at least 80% and a specificity of at least 80% could be achieved. A test with reasonable (at least 80%) sensitivity and specificity (Consensus report of the Working Group on:

"Molecular and Biochemical Markers of Alzheimer's Disease". The Ronald and Nancy Reagan Research Institute of the Alzheimer's Association and the National Institute on Aging Working Group. Neurobiol Aging. Mar-Apr 1998; 19(2): 109-116) could be useful as a screening tool to identify individuals at risk who would then be eligible for more confirmatory cognitive, CSF and imaging clinical trial enrollment testing.

Biomarkers to Detect Alzheimer's Disease

A biomarker is an organic biomolecule which is differentially present in a sample taken from an individual of one phenotypic status (e.g., having a disease) as compared with an individual of another phenotypic status (e.g., not having the disease).

A biomarker is differentially present between the two individuals if the mean or median expression level of the biomarker in the different individuals is calculated to be statistically significant. Biomarkers, alone or in combination, provide measures of relative risk that an individual belongs to one phenotypic status or another. Therefore, they are useful as markers for diagnosis of disease, therapeutic effectiveness of a drug, and drug toxicity.

Accordingly, the invention provides methods for identifying one or more biomarkers which can be used to aid in the diagnosis, to diagnose, detect, predict neurological diseases such as neurodegenerative disorders. The methods of the invention are carried out by obtaining a set of measured values for a plurality of biomarkers from a biological sample derived from a test individual, obtaining a set of measured values for a plurality of biomarkers from a biological sample derived from a control individual, comparing the measured values for each biomarker between the test and control sample, and identifying biomarkers which are significantly different between the test value and the control value also referred to as a reference value.

The process of comparing a measured value and a reference value can be carried out in any convenient manner appropriate to the type of measured value and reference value for the AD biomarker at issue. For example, "measuring" can be performed using quantitative or qualitative measurement techniques, and the mode of comparing a measured value and a reference value can vary depending on the

measurement technology employed. For example, when a qualitative calorimetric assay is used to measure AD biomarker levels, the levels may be compared by visually comparing the intensity of the colored reaction product, or by comparing data from densitometric or spectrometric measurements of the colored reaction product (e.g., comparing numerical data or graphical data, such as bar charts, derived from the measuring device). However, it is expected that the measured values used in the methods of the invention will most commonly be quantitative values (e.g., quantitative

measurements of concentration. In other examples, measured values are qualitative. As with qualitative measurements, the comparison can be made by inspecting the numerical data, by inspecting representations of the data (e.g., inspecting graphical representations such as bar or line graphs).

A measured value is generally considered to be substantially equal to or greater than a reference value if it is at least about 95% of the value of the reference. A measured value is considered less than a reference value if the measured value is less than about 95% of the reference value. A measured value is considered more than a reference value if the measured value is at least more than about 5% greater than the reference value.

The process of comparing may be manual (such as visual inspection by the practitioner of the method) or it may be automated. For example, an assay device (such as a luminometer for measuring chemiluminescent signals) may include circuitry and software enabling it to compare a measured value with a reference value for an AD biomarker. Alternately, a separate device (e.g., a digital computer) may be used to compare the measured value(s) and the reference value(s). Automated devices for comparison may include stored reference values for the AD biomarker(s) being measured, or they may compare the measured value(s) with reference values that are derived from contemporaneously measured reference samples. In some embodiments, the methods of the invention utilize "simple" or "binary" comparison between the measured level(s) and the reference level(s) (e.g., the comparison between a measured level and a reference level determines whether the measured level is higher or lower than the reference level). For AD biomarkers, a comparison showing that the measured value for the biomarker is altered from the reference value indicates or suggests a diagnosis of AD.

By way of a non-limiting example, the process of comparing the measured values to identify indicators for disease state can be carried out using multiple marker analysis approach including but not limited to Logistic Regression, Boosted Tree Models, Flexible Discriminant Analysis (FDA), K-Nearest Neighbors (KNN), Naive Bayes,

Partial Least Squares (PLS), Random Forests, Shrunken Centroids, Sparse Partial Least Squares and Support Vector Machines approaches.

One aspect of the present invention provides biomarkers to detect Alzheimer's disease. A biomarker is typically a protein, found in a bodily fluid, whose level varies with disease state and may be readily quantified. The quantified level may then be compared to a known value. The comparison may be used for several different purposes, including but not limited to, diagnosis of Alzheimer's disease, prognosis of Alzheimer's disease, and monitoring treatment of Alzheimer's disease.

Through screening performed as detailed in the examples provided elsewhere herein, several biomarkers have been identified as associated with Alzheimer's disease. Examples of biomarkers identified as being indicators of Alzheimer's disease include, but is not limited, to Alpha- 1 Microglobulin, Angiopoietin-2, Apolipoprotein E, Beta-2 Microglobulin, BLC, Cortisol, E-Selectin, FAS, Fatty Acid Binding Protein, Hepatocyte Growth Factor, IGF BP-2, Interleukin 10, NT-Pro-BNP, Osteopontin, Pancreatic Polypeptide, PAPP-A, Resistin, Stem Cell Factor, Tenascin C,

Thrombomodulin, TIMP-1, VCAM-1, VEGF, Von Willebrand Factor, and any combination thereof. In some instances, age, gender, and ApoE genotype (e2/e2, e2/e3, e2/e4, e3/e3, e3/e4 and e4/e4) are additional factors that are considered in evaluating the diagnosis of an individual for Alzheimer's disease.

In one aspect, the invention provides compositions for use in the methods of the invention. In one embodiment, the composition comprises a reagent that detects and/or quantitates an Alzheimer Disease biomarker. In some embodiments, the composition comprises a panel of reagents, each of which detects and/or quantitates a different Alzheimer Disease biomarker. In some embodiments, the Alzheimer Disease biomarker is an Alzheimer Disease biomarker shown in Table 2. In some embodiments, the composition comprises two or more reagents, each of which detects and/or quantitates a different one of two or more, three or more, four or more or five or more Alzheimer Disease biomarkers. In some embodiments, the compositions comprises a panel of reagents each of which detects and/or quantitates a different one of 24 Alzheimer Disease biomarkers shown in Table 2. In some embodiments, the composition comprises a panel of reagents for the detection and/or quantification of Cortisol, pancreatic polypeptide, osteopontin, IGF BP2, and resistin. In some embodiments, the composition comprises a panel of reagents for the detection and/or quantification of resistin, pancreatic polypeptide, ApoD, G-CSF, MlPlbeta, ApoE, NrCAM, MMP2, Cortisol, PDGF-BB,

MMP 1, and 1-309. In some embodiments, the composition comprises a panel of reagents for the detection and/or quantification of Alpha- 1 Microglobulin, Angiopoietin-2, Apolipoprotein E, Beta-2 Microglobulin, Chemokine (C-X-C motif) ligand 13 (CXCL13) also known as B lymphocyte chemoattractant (BLC), Cortisol, E-Selectin, FAS, Fatty Acid Binding Protein, Hepatocyte Growth Factor, IGF BP-2, Interleukin 10, NT-Pro-BNP (Brain natriuretic peptide co-secreted along with a 76 amino acid N-terminal fragment), Osteopontin, Pancreatic Polypeptide, PAPP-A, Resistin, Stem Cell Factor, Tenascin C, Thrombomodulin, tissue inhibitor of metalloproteinases-1 (TIMP-1), vascular cell adhesion molecule- 1 (VCAM-1), Vascular endothelial growth factor (VEGF), and Von Willebrand Factor.

In one embodiment, the combination of biomarkers associated with assessing Alzheimer's disease is collectively presented on a detectable medium referred to as a panel or plasma based panel. Each of the biomarkers identified above may be used in concert with other biomarkers. For instance, 2 or more, 3 or more, 4 or more, 5 or more, 6 or more, 7 or more, 8 or more, 10 or more, 15 or more, 20 or more, 30 or more, 40 or more, (and any integer value in between) Alzheimer's disease biomarkers may be used in concert with other, and with other known or to be known biomarkers for

Alzheimer's disease.

In some embodiments, the composition is an array of two or more detection reagents, e.g., antibodies and/or oligonucleotides each of which binds to a different one of two Alzheimer Disease biomarker proteins or nucleic acids, respectively. In a preferred embodiment, the Alzheimer Disease biomarkers are proteins and the array contains antibodies sufficient to measure a statistically significant alteration in Alzheimer Disease biomarker expression compared to a reference value. In some embodiments, the invention provides a kit with at least two detection reagents, each of which binds to a different one of two Alzheimer Disease biomarker proteins, metabolites, or other analytes.

In some embodiments, a. detection reagent is immobilized on a solid matrix such as a porous strip or bead to form at least one Alzheimer Disease biomarker detection site,

Provided herein are methods for assessing cognitive function, assessing cognitive impairment, diagnosing or aiding diagnosis of cognitive impairment by obtaining measured levels of one or more AD diagnosis biomarkers in a biological sample from an individual, such as for example, a plasma sample from an individual, and comparing those measured levels to reference levels.

The levels of Alzheimer's disease biomarkers of the invention may be assessed in several different biological samples, for example bodily fluids. Non-limiting examples of bodily fluid include whole blood, plasma, serum, bile, lymph, pleural fluid, semen, saliva, sweat, urine, and CSF. In one embodiment, the bodily fluid is selected from the group of whole blood, plasma, and serum. In another embodiment, the bodily fluid is whole blood. In yet another embodiment, the bodily fluid is plasma. In still yet another embodiment, the bodily fluid is serum.

The bodily fluid is obtained from the individual using conventional methods in the art. For instance, one skilled in the art knows how to draw blood and how to process it in order to obtain serum and/or plasma for use in the method. Generally speaking, the method preferably maintains the integrity of the biomarkers of the invention such that it can be accurately quantified in the bodily fluid. Methods for collecting blood or fractions thereof are well known in the art. For example, see US Patent No. 5,286,262, which is hereby incorporated by reference in its entirety.

A bodily fluid may be obtained from any mammal known to suffer from Alzheimer's disease or that can be used as a disease model for Alzheimer's disease. In one embodiment, the mammal is a rodent. Examples of rodents include mice, rats, and guinea pigs. In another embodiment, the mammal is a primate. Examples of primates include monkeys, apes, and humans. In an exemplary embodiment, the mammal is a human. In some embodiments, the individual has no clinical signs of Alzheimer's disease. In other embodiments, the individual has mild clinical signs of Alzheimer's disease. In yet other embodiments, the individual may be at risk for Alzheimer's disease. In still other embodiments, the individual has been diagnosed with Alzheimer's disease. Assessment of biomarker levels may encompass assessment of the level of protein concentration or the level of enzymatic activity of the biomarker, wherever applies. In either case, the level is quantified such that a value, an average value, or a range of values is determined. In one embodiment, the level of protein concentration of the Alzheimer's disease biomarker is quantified.

There are numerous known methods and kits for measuring the amount or concentration of a protein in a sample, including as non-limiting examples, ELISA, western blot, absorption measurement, colorimethc determination, Lowry assay, Bicinchoninic acid assay, or a Bradford assay. Commercial kits include ProteoQwest™ Colohmetric Western Blotting Kits (Sigma-Aldrich, Co.), QuantiPro™ bicinchoninic acid (BCA) Protein Assay Kit (Sigma-Aldrich, Co.), FluoroProfile™ Protein Quantification Kit (Sigma-Aldrich, Co.), the Coomassie Plus - The Better Bradford Assay (Pierce Biotechnology, Inc.), and the Modified Lowry Protein Assay Kit (Pierce Biotechnology, Inc.). In certain embodiments, the protein concentration is measured using a luminex based multiplex immunoassay panel. However, the invention should not be limited to any particular assay for assessing the level of a biomarker of the invention. That is, any currently known assay used to detect protein levels and assays to be discovered in the future can be used to detect the biomarkers of the invention.

Methods of quantitatively assessing the level of a protein in a biological sample such as plasma are well known in the art. In some embodiments, assessing the level of a protein involves the use of a detector molecule for the biomarker. Detector molecules can be obtained from commercial vendors or can be prepared using conventional methods available in the art. Exemplary detector molecules include, but are not limited to, an antibody that binds specifically to the biomarker, a naturally-occurring cognate receptor, or functional domain thereof, for the biomarker, and a small molecule that binds specifically to the biomarker.

In a preferred embodiment, the level of a biomarker is assessed using an antibody. Thus, non-limiting exemplary methods for assessing the level of a biomarker in a biological sample include various immunoassays, for example, immunohistochemistry assays, immunocytochemistry assays, ELISA, capture ELISA, sandwich assays, enzyme immunoassay, radioimmunoassay, fluorescent immunoassay, and the like, all of which are known to those of skill in the art. See e.g. Harlow et ah, 1988, Antibodies: A Laboratory Manual, Cold Spring Harbor, New York; Harlow et ah, 1999, Using

Antibodies: A Laboratory Manual, Cold Spring Harbor Laboratory Press, NY. The generation of polyclonal antibodies is accomplished by inoculating the desired animal with an antigen and isolating antibodies which specifically bind the antigen therefrom.

Monoclonal antibodies directed against one of the biomarkers identified herein may be prepared using any well known monoclonal antibody preparation procedures, such as those described, for example, in Harlow et al. (1988, In: Antibodies, A Laboratory Manual, Cold Spring Harbor, NY) and in Tuszynski et al. (1988, Blood, 72: 109-1 15). Human monoclonal antibodies may be prepared by the method described in U.S. patent publication 2003/0224490. Monoclonal antibodies directed against a biomarker such as EGF are generated from mice immunized with the biomarker using standard procedures as referenced herein.

For use in preparing an antibody, a biomarker may be purified from a biological source that endogenously comprises the biomarker, or from a biological source recombinantly-engineered to produce or over-produce the biomarker, using conventional methods known in the art. Exemplary nucleic acid and protein sequences for the eleven biomarkers described herein are readily available in public sequence databases, such as National Library of Medicine's genetic sequence database GenBank® (Benson et al, 2008, Nucleic Acids Research, 36(Database issue):D25-30). Preferably, antibodies are generated against the human homologs of each of the eleven biomarkers for practicing the methods using a biological sample from a human individual diagnosed with Parkinson's Disease.

Nucleic acid encoding the monoclonal antibody obtained using the procedures described herein may be cloned and sequenced using technology which is available in the art, and is described, for example, in Wright et al. (1992, Critical Rev. Immunol. 12(3,4): 125-168) and the references cited therein. Further, the antibody useful in the practice of the invention may be "humanized" using the technology described in Wright et al, (supra) and in the references cited therein, and in Gu et al. (1997,

Thrombosis and Hematocyst 77(4):755-759).

To generate a phage antibody library, a cDNA library is first obtained from mRNA which is isolated from cells, e.g., the hybridoma, which express the desired protein to be expressed on the phage surface, e.g., the desired antibody. cDNA copies of the mRNA are produced using reverse transcriptase. cDNA which specifies

immunoglobulin fragments are obtained by PCR and the resulting DNA is cloned into a suitable bacteriophage vector to generate a bacteriophage DNA library comprising DNA specifying immunoglobulin genes. The procedures for making a bacteriophage library comprising heterologous DNA are well known in the art and are described, for example, in Sambrook et al. (2001, Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY).

Bacteriophage which encode the desired antibody may be engineered such that the protein is displayed on the surface thereof in such a manner that it is available for binding to the antigen against which the antibody is directed. Thus, when bacteriophage which express a specific antibody are incubated in the presence of the antigen, for instance, antigen immobilized on a resin or surface, the bacteriophage will bind to the antigen. Bacteriophage which do not express the antibody will not bind to the antigen. Such panning techniques are well known in the art and are described for example, in Wright et al, (supra).

Processes, such as those described above, have been developed for the production of human antibodies using M13 bacteriophage display (Burton et al, 1994, Adv. Immunol. 57: 191-280). Essentially, a cDNA library is generated from mRNA obtained from a population of antibody -producing cells. The mRNA encodes rearranged immunoglobulin genes and thus, the cDNA encodes the same. Amplified cDNA is cloned into M13 expression vectors creating a library of phage which express human Fab fragments on their surface. Phage which display the antibody of interest are selected by antigen binding and are propagated in bacteria to produce soluble human Fab

immunoglobulin. Thus, in contrast to conventional monoclonal antibody synthesis, this procedure immortalizes DNA encoding human immunoglobulin rather than cells which express human immunoglobulin.

The procedures just presented describe the generation of phage which encode the Fab portion of an antibody molecule. However, phage which encode single chain antibodies (scFv/phage antibody libraries) are also useful in preparing Fab molecules useful in the invention. Fab molecules comprise the entire Ig light chain, that is, they comprise both the variable and constant region of the light chain, but include only the variable region and first constant region domain (CHI) of the heavy chain. Single chain antibody molecules comprise a single chain of protein comprising the Ig Fv fragment. An Ig Fv fragment includes only the variable regions of the heavy and light chains of the antibody, having no constant region contained therein. Phage libraries comprising scFv DNA may be generated following the procedures described in Marks et al, 1991, J. Mol. Biol. 222:581-597. Panning of phage so generated for the isolation of a desired antibody is conducted in a manner similar to that described for phage libraries comprising Fab DNA. Synthetic phage display libraries in which the heavy and light chain variable regions may be synthesized such that they include nearly all possible specificities (Barbas, 1995, Nature Medicine 1 :837-839; de Kruif et al, 1995, J. Mol. Biol. 248:97-105) may also be used to prepare an antibody useful in the practice of the invention.

Other methods for assessing the level of a protein include chromatography (e.g., HPLC, gas chromatography, liquid chromatography) and mass spectrometry (e.g., MS, MS-MS). For instance, a chromatography medium comprising a cognate receptor for the biomarker or a small molecule that binds to the biomarker can be used to substantially isolate the biomarker from the biological sample. Small molecules that bind specifically to a biomarker can be identified using conventional methods in the art, for instance, screening of compounds using combinatorial library methods known in the art, including biological libraries, spatially-addressable parallel solid phase or solution phase libraries, synthetic library methods requiring deconvolution, the "one-bead one- compound" library method, and synthetic library methods using affinity chromatography selection.

The level of substantially isolated protein can be quantitated directly or indirectly using a conventional technique in the art such as spectrometry, Bradford protein assay, Lowry protein assay, biuret protein assay, or bicinchoninic acid protein assay, as well as immunodetection methods.

In another embodiment, the level of enzymatic activity of the biomarker if such biomarker has an enzymatic activity maybe quantified. Generally, enzyme activity may be measured by means known in the art, such as measurement of product formation, substrate degradation, or substrate concentration, at a selected point(s) or time(s) in the enzymatic reaction. There are numerous known methods and kits for measuring enzyme activity. For example, see US Patent No. 5,654, 152. Some methods may require purification of the Alzheimer's disease biomarker prior to measuring the enzymatic activity of the biomarker. A pure biomarker constitutes at least about 90%, preferably, 95% and even more preferably, at least about 99% by weight of the total protein in a given sample. Alzheimer's disease biomarkers of the invention may be purified according to methods known in the art, including, but not limited to, ion-exchange chromatography, size-exclusion chromatography, affinity chromatography, differential solubility, differential centrifugation, and HPLC. Determination of the Status of Alzheimer's Disease

The present invention is based on biomarker profiles or signatures determined for biological samples from individuals diagnosed with Alzheimer's Disease as well as from one or more other groups of control individuals (e.g., healthy control subjects not diagnosed with Alzheimer's Disease). The profile for Alzheimer's Disease was compared to the profile for biological samples from the one or more other groups of control individual. Those molecules differentially present, including those molecules differentially present at a level that is statistically significant, in the profile of Alzheimer's Disease samples as compared to another group (e.g., healthy control subjects not diagnosed with Alzheimer's Disease) were identified as biomarkers to distinguish those groups. The biomarkers are discussed in more detail elsewhere herein.

As described herein, assessment of results can depend on the statistical models used to predict the data. For example, AD biomarker levels relative to another reference level, which may be relative to the level of another AD biomarker, may be obtained according to an appropriate statistical model disclosed herein. In other methods described herein, high correlation amongst many biomarkers was observed to provide an indicator of disease state when analyzed using an appropriate statistical model.

The reference level used for comparison with the measured level for a AD biomarker may vary, depending on aspect of the invention being practiced, as will be understood from the foregoing discussion. For detection of AD, the "reference level" is typically a predetermined reference level, such as an average of levels obtained from a population that is not afflicted with AD, but in some instances, the reference level can be a mean or median level from a group of individuals including AD patients. In some instances, the predetermined reference level is derived from (e.g., is the mean or median of) levels obtained from an age-matched population. In some instances, the age-matched population comprises individuals with non-AD neurodegenerative disorders. In some instances, the reference level may be a historical reference level for the particular patient (e.g., a biomarker level that was obtained from a sample derived from the same individual, but at an earlier point in time). In some instances, the predetermined reference level is derived from (e.g., is the mean or median of) levels obtained from an age-matched population.

Age-matched populations (from which reference values may be obtained) are ideally the same age as the individual being tested, but approximately age-matched populations are also acceptable. Approximately age-matched populations may be within 1, 2, 3, 4, or 5 years of the age of the individual tested, or may be groups of different ages which encompass the age of the individual being tested. Approximately age-matched populations may be in 2, 3, 4, 5, 6, 7, 8, 9, or year increments (e.g. a "5 year increment" group which serves as the source for reference values for a 62 year old individual might include 58-62 year old individuals, 59-63 year old individuals, 60-64 year old individuals, 61-65 year old individuals, or 62-66 year old individuals.

The level(s) of the one or more biomarkers may be compared to

Alzheimer's Disease-positive and/or Alzheimer's Disease-negative reference levels using various techniques, including a simple comparison (e.g., a manual comparison) of the level(s) of the one or more biomarkers in the biological sample to Alzheimer's Disease- positive and/or Alzheimer's Disease-negative reference levels. The level(s) of the one or more biomarkers in the biological sample may also be compared to Alzheimer's Disease- positive and/or Alzheimer's Disease-negative reference levels using one or more statistical analyses.

Statistical models useful in the present invention includes but are not limited to Logistic Regression, Boosted Tree Models, Flexible Discriminant Analysis (FDA), K-Nearest Neighbors (KNN), Na^'ive Bayes, Partial Least Squares (PLS), Random Forests, Shrunken Centroids, Sparse Partial Least Squares and Support Vector Machines approaches. Preferably, the statistical model is Partial Least Squares.

In some instances, the AD biomarkers of the invention include one or more of the biomarkers shown in Table 2. In one embodiment, the combination of biomarkers associated with assessing Alzheimer's disease is collectively presented on a detectable medium referred to as a panel or plasma based panel. Each of the biomarkers identified above may be used in concert with other biomarkers. For instance, 2 or more, 3 or more, 4 or more, 5 or more, 6 or more, 7 or more, 8 or more, 10 or more, 15 or more, 20 or more, 30 or more, 40 or more, (and any integer value in between) Alzheimer's disease biomarkers may be used in concert with other, and with other known or to be known biomarkers for Alzheimer's disease.

In some instances, age, gender, and ApoE genotype (e2/e2, e2/e3, e2/e4, e3/e3, e3/e4 and e4/e4) are additional factors that are considered in identifying an individual for Alzheimer's disease. Therefore, in some instances, predictors for AD in an individual are selected from the group consisting of Alpha- 1 Microglobulin,

Angiopoietin-2, Apolipoprotein E, Beta-2 Microglobulin, BLC, Cortisol, E-Selectin, FAS, Fatty Acid Binding Protein, Hepatocyte Growth Factor, IGF BP-2, Interleukin 10, NT-Pro-BNP, Osteopontin, Pancreatic Polypeptide, PAPP-A, Resistin, Stem Cell Factor, Tenascin C, Thrombomodulin, TIMP-1, VCAM-1, VEGF, Von Willebrand Factor, age, gender, ApoE genotype (e2/e2, e2/e3, e2/e4, e3/e3, e3/e4 and e4/e4), and any combination thereof.

In other instances, predictors of AD in an individual are selected from the group consisting of Resistin, e3/e3, Pancreatic Polypeptide, e3/e4, ApoD, G-CSF, MlPlbeta, ApoE, NrCAM, MMP2, Cortisol, PDGF-BB, e2/e3, MMP 1, 1-309, and any combination thereof.

In other instances, predictors of AD in an individual are selected from the group consisting of Cortisol, Pancreatic Polypeptide, osteopontin, IGF BP2, Resistin, and any combination thereof.

Based on the disclosure presented herein, a skilled artisan would understand that a plasma based signature for Alzheimer's disease can be detected in plasma and that signature can differentiate AD from healthy controls and other forms of dementia. The signature for Alzheimer's disease includes the combination of biomarkers disclosed herein. In some instances, the signature for Alzheimer's disease is a combination of biomarkers and predictors of Alzheimer's disease disclosed herein. For example, the biomarkers of the invention in combination with other factors such as age, gender, ApoE genotype (e2/e2, e2/e3, e2/e4, e3/e3, e3/e4 and e4/e4), can improve diagnostic and screening accuracy. The biomarkers can also be combined with cognitive tests such as a simple memory test to improve diagnostic and screening accuracy. In some instances, the biomarkers of the invention can be combined with additional confirmatory CSF and imaging testing.

Additionally, Tables 2 provide a listing of biomarkers that are useful for identifying an individual with AD. In some instances, certain biomarkers are increases and other biomarkers are decreased in AD compared to age-matched normal controls (Tables 10, 11, and 12. Generally, a significant increase in a biomarker as compared to an appropriate control is indicative of AD, and a significant decrease in a biomarker as compared to an appropriate control is indicative of AD. In some instances, correlation between biomarkers provide an indication of AD in an individual. In some examples, any one or more of the biomarkers listed in Table 2 can be used to identify AD in an individual as distinguished from other non-AD individuals. The biomarkers of the invention can be used in diagnostic tests to assess the status of Alzheimer's disease in an individual, e.g., to diagnose Alzheimer's disease or to assess the degree of Alzheimer's disease in the individual. The phrase "Alzheimer's disease status" includes any distinguishable manifestation of the disease, including non- Alzheimer's disease, e.g., normal or non-demented. For example, disease status includes, without limitation, the presence or absence of Alzheimer's disease (e.g., Alzheimer's disease v. non-Alzheimer's disease), the risk of developing disease, the stage of the disease, the progress of disease (e.g., progress of disease or remission of disease over time) and the effectiveness or response to treatment of disease. Based on this status, further procedures may be indicated, including additional diagnostic tests or therapeutic procedures or regimens.

The ability of a diagnostic test to correctly predict status is commonly measured based on the sensitivity of the assay, the specificity of the assay or the area under a receiver operated characteristic ("ROC") curve. Sensitivity is the percentage of true positives that are predicted by a test to be positive, while specificity is the percentage of true negatives that are predicted by a test to be negative. An ROC curve provides the sensitivity of a test. The greater the area under the ROC curve, the more powerful the predictive value of the test. Other useful measures of the utility of a test are positive predictive value and negative predictive value. Positive predictive value is the percentage of people who test positive that is actually positive. Negative predictive value is the percentage of people who test negative that is actually negative.

As apparent from the example disclosed herein, diagnostic tests that use the biomarkers of the invention exhibit a sensitivity and specificity of at least 75%, at least 80%, at least 85%, at least 90%, at least 95%, at least 98% and about 100%. In some instances, screening tools of the present invention exhibits a high sensitivity of at least 75%, at least 80%, at least 85%, at least 90%, at least 95%, at least 98% and about 100%. Without wishing to be bound by any particular theory, it is believed that screening tools should exhibit high sensitivity, but specificity can be low. However, Diagnostics should have high sensitivity and specificity.

While individual biomarkers are useful diagnostic biomarkers, it has been found that a combination of biomarkers can provide greater predictive value of a particular status than single biomarkers alone. Specifically, the detection of a plurality of biomarkers in a sample can increase the sensitivity and/or specificity of the test. A combination of at least two biomarkers is sometimes referred to as a "biomarker profile" or "biomarker fingerprint." Examples of Alzheimer's disease biomarkers of the invention include, but are not limited, to Alpha- 1 Microglobulin, Angiopoietin-2, Apolipoprotein E, Beta-2 Microglobulin, BLC, Cortisol, E-Selectin, FAS, Fatty Acid Binding Protein, Hepatocyte Growth Factor, IGF BP-2, Interleukin 10, NT-Pro-BNP, Osteopontin, Pancreatic Polypeptide, PAPP-A, Resistin, Stem Cell Factor, Tenascin C,

In addition, the methods disclosed herein using the biomarkers listed in the tables presented herein may be used in combination with clinical diagnostic measures of Alzheimer's Disease and/or other neurodegenerative diseases. Combinations with clinical diagnostics may facilitate the disclosed methods, or confirm results of the disclosed methods (for example, facilitating or confirming diagnosis, monitoring progression or regression, and/or determining predisposition to Alzheimer's Disease).

Determining Alzheimer's disease status typically involves classifying an individual into one of two or more groups based on the results of the diagnostic test. The diagnostic tests described herein can be used to classify an individual into a number of different states. In one embodiment, the invention provides methods for determining the presence or absence of Alzheimer's disease in an individual (status: Alzheimer's disease v. non-Alzheimer's disease). The presence or absence of Alzheimer's disease is determined by measuring the relevant biomarker or biomarkers in samples obtained from individuals and then either submitting them to a classification algorithm or comparing them with a reference amount and/or pattern of biomarkers that is associated with the particular risk level.

In another embodiment, the invention provides methods for determining the risk of developing disease in an individual. Biomarker amounts or patterns are characteristic of various risk states, e.g., high, medium or low. The risk of developing Alzheimer's disease is determined by measuring the relevant biomarker or biomarkers in sample obtained from individuals and then either submitting them to a classification algorithm or comparing them with a reference amount and/or pattern of biomarkers that is associated with the particular risk level.

In yet another embodiment, the invention provides methods for determining the stage of Alzheimer's disease in an individual. Each stage of the disease can be characterized by the amount of a biomarker or relative amounts of a set of biomarkers (i.e., a pattern) that are found in a sample obtained from the individual. The stage of Alzheimer's disease is determined by measuring the relevant biomarker or biomarkers and then either submitting them to a classification algorithm or comparing them with a reference amount and/or pattern of biomarkers that is associated with the particular stage.

In another embodiment, the invention provides methods for determining the course of Alzheimer's disease in an individual. Disease course refers to changes in disease status over time, including disease progression (worsening) and disease regression (improvement). Over time, the amounts or relative amounts (e.g., the pattern) of the biomarkers changes. For example, levels of various biomarkers of the present invention increase with progression of disease. Accordingly, this method involves measuring the level of one or more biomarkers in an individual at two or more different time points, e.g., a first time and a second time, and comparing the change in amounts. The course of disease is determined based on these comparisons.

In some instances, the levels of various biomarkers of the invention decreases with disease progression. In this method, the level of one or more biomarkers in a sample from an individual is measured at two or more different time points, e.g., a first time and a second time, and the change in levels, if any is assessed. The course of disease is determined based on these comparisons.

Similarly, changes in the rate of disease progression (or regression) may be monitored by measuring the level of one or more biomarkers at different times and calculating the rate of change in biomarker levels. The ability to measure disease state or rate of disease progression is important for drug treatment studies where the goal is to slow down or arrest disease progression using therapy.

Additional embodiments of the invention relate to the communication of the results or diagnoses or both to technicians, physicians or patients, for example. In certain embodiments, computers are used to communicate results or diagnoses or both to interested parties, e.g., physicians and their patients.

In certain embodiments, the methods of the invention further comprise managing individual treatment based on their disease status. Such management includes the actions of the physician or clinician subsequent to determining Alzheimer's disease status. For example, if a physician makes a diagnosis of Alzheimer's disease, then a certain regime of treatment, such as prescription or administration of the therapeutic compound might follow. Alternatively, a diagnosis of non-Alzheimer's disease might be followed by further testing to determine any other diseases that might the patient might be suffering from. Also, if the test is inconclusive with respect to Alzheimer's disease status, further tests may be called for.

In a preferred embodiment of the invention, a diagnosis based on the presence or absence or relative levels in the biological sample of an individual of the relevant biomarkers disclosed herein is communicated to the individual as soon as possible after the diagnosis is obtained.

According to yet another aspect, the present invention provides a method of assessing efficacy of a treatment of Alzheimer's disease in a patient comprising: a) determining a baseline level of biomarkers in a first sample obtained from the patient before receiving the treatment; b) determining the level of same biomarkers in a second sample obtained from the patient after receiving the treatment; wherein an alteration in the levels of the biomarkers in the post-treatment sample is correlated with a positive treatment outcome.

Assays

The experiments disclosed herein were designed to develop an assay to identify universal accepted biomarkers for diagnosing, screening, monitoring and staging neurodegenerative diseases such as Alzheimer's disease that are fast, more accurate, and less expensive. The results presented herein demonstrate that a diagnostic assay has been developed that can detect among other things early onset of Alzheimer's disease.

Detection of early onset of Alzheimer's disease is believed to increase the success rate of the individual being successfully treated for Alzheimer's disease.

The diagnostic method of the present invention can be applied to subjects who have been previously diagnosed with Alzheimer's disease, those who are suspected of having Alzheimer's disease, and those at risk of developing Alzheimer's disease. For example, patients diagnosed with dementia, in particular, those patients who were previously clinically normal, are suitable subjects. However, it is not intended that the present invention be limited to use with any particular subject types.

According to some embodiments, the subject is a human subject.

According to certain embodiments, the subject is selected from the group consisting of subjects displaying pathology resulting from Alzheimer's disease, subjects suspected of displaying pathology resulting from Alzheimer's disease, and subjects at risk of displaying pathology resulting from Alzheimer's disease.

According to another embodiment, the Alzheimer's disease diagnosed using the method of the present invention is selected from the group consisting of late onset Alzheimer's disease, early onset Alzheimer's disease, familial Alzheimer's disease and sporadic Alzheimer's disease.

Early-onset Alzheimer's disease (EOAD) is a rare form of Alzheimer's disease in which individuals are diagnosed with the disease before age 65. Less than 10% of all Alzheimer's disease patients have EOAD. Younger individuals who develop Alzheimer's disease exhibit more of the brain abnormalities that are normally associated with Alzheimer's disease. EOAD is usually familial and follows an autosomal dominant inheritance pattern. To date, mutations in three genes including amyloid precursor protein (APP) on chromosome 21, presenilin 1 (PSEN1) on chromosome 14 and presenilin 2 (PSEN2) on chromosome 1 have been identified in families with EOAD . Mutations in the APP, PSEN1 and PSEN2 genes account for about 50% of the disease . Most of the pathogenic mutations in the APP and presenilin genes are associated with abnormal processing of APP, which leads to the overproduction of toxic Αβ-1-42. Down syndrome patients, who have three copies of chromosome 21 which includes the APP gene, begin to develop the characteristic senile plaques and tau tangles at the ages of 30 and 40 (M. Ilyas Kamboh (2004), Molecular Genetics of Late-Onset Alzheimer's Disease, Annals of Human Genetics 68(4):381-404).

Late-onset Alzheimer's disease (LOAD) is the most common form of Alzheimer's disease, accounting for about 90% of cases and usually occurring after age 65. LOAD strikes almost half of all individuals over the age of 85 and may or may not be hereditary. It is a complex and multifactorial disease with the possible involvement of several genes. Genome-wide linkage or linkage disequilibrium studies on LOAD have provided informative data for the existence of multiple putative genes for Alzheimer's disease on several chromosomes, with the strongest evidence on chromosomes 12, 10, 9 and 6. LOAD cases tend to be sporadic, wherein there is no family history of the disease. Genetic susceptibility at multiple genes and interaction between these genes as well as environmental factors are most likely responsible for the etiology of LOAD. Twin data on incident cases indicates that almost 80% of the LOAD risk is attributable to genetic factors. The Apolipoprotein E (APOE) gene on chromosome 19ql3 has been identified as a strong risk factor for LOAD. In fact, the ΑΡΟΕ-ε4 allele has been established as a strong susceptibility marker that accounts for nearly 30% of the risk in late-onset Alzheimer's disease. More specifically, three variants of APOE, encoded by codons 1 12 and 158, have been found to modify the risk of LOAD. As compared to the common APOE-83 allele (codon 1 12=Cys and codon 158=Arg), the ΑΡΟΕ-ε4 allele (codon 112=Arg and codon 158=Arg) increases the risk of Alzheimer's disease, while the ΑΡΟΕ-ε2 allele (codon 1 12=Cys and codon 158=Cys) decreases the risk of Alzheimer's disease. The effect of the ΑΡΟΕ-ε4 allele is dose related, wherein one or two copies of the ΑΡΟΕ-ε4 allele are associated with 3-fold or 15-fold risk, respectively. However, the effect of the ΑΡΟΕ-ε4 allele on Alzheimer's disease risk appears to decline with increasing age (M. Ilyas Kamboh (2004), supra).

Based on the disclosure presented herein, a skilled artisan would understand that a profile of biomarkers for Alzheimer's disease can be detected in plasma and that profile can differentiate AD from healthy controls and other forms of dementia. The profiles for Alzheimer's disease includes the biomarkers disclosed herein. In some instances, the profile for Alzheimer's disease is a combination of biomarkers and other factors of Alzheimer's disease disclosed herein. For example, the biomarkers of the invention in combination with other factors such as age, gender, ApoE genotype (e2/e2, e2/e3, e2/e4, e3/e3, e3/e4 and e4/e4), can improve diagnostic and screening accuracy. The biomarkers can also be combined with cognitive tests such as a simple memory test to improve diagnostic and screening accuracy. In some instances, the biomarkers of the invention can be combined with additional confirmatory CSF and imaging testing.

For example, the biomarkers of the invention can be combined with existing criteria for dementia to improve diagnostic and screening accuracy of

Alzheimer's disease. Dementia is the decline of memory and other cognitive functions in comparison with the patient's previous level of function as determined by a history of decline in performance and by abnormalities noted from clinical examination and neuropsychological tests. A diagnosis of dementia cannot be made when consciousness is impaired by delirium, drowsiness, stupor, or coma or when other clinical abnormalities prevent adequate evaluation of mental status. Dementia is a diagnosis based on behavior and cannot be determined by computerized tomography, electroencephalography, or other laboratory instructions, although specific causes of dementia may be identified by these means. In some instances, the biomarkers of the invention can be combined with existing criteria Alzheimer's disease. A clinical diagnosis of probable Alzheimer's disease can be made with confidence if there is a typical insidious onset of dementia with progression and if there are no other systemic or brain diseases that could account for the progressive memory and other cognitive deficits. Among the disorders that must be excluded are manic depressive disorder, Parkinson's disease, multiinfarct dementia, and drug intoxication; less commonly encountered disorders that may cause dementia include thyroid disease, pernicious anemia, luetic brain disease and other chronic infections of the nervous system, subdural hematoma, occult hydrocephalus, Huntington's disease, Creutzfeldt- Jakob disease, and brain tumors.

A diagnosis of definite Alzheimer's disease requires histopathologic confirmation. A clinical diagnosis of possible Alzheimer's disease may be made in the presence of other significant diseases, particularly if, on clinical judgment, Alzheimer's disease is considered the more likely cause of the progressive dementia. The clinical diagnosis of possible rather than probable Alzheimer's disease may be used if the presentation or course is somewhat aberrant. The information needed to apply these criteria is obtained by standard methods of examination: the medical history; neurologic; psychiatric, and clinical examinations; neuropsychological tests; and laboratory studies.

It is believed that some of the biomarkers for Alzheimer's Disease described herein may also be biomarkers for neurodegenerative diseases in general.

Therefore, it is believed that at least some of the Alzheimer's Disease biomarkers may be used in the methods described herein for neurodegenerative diseases in general. That is, the methods described herein with respect to Alzheimer's Disease may also be used for diagnosing (or aiding in the diagnosis of) a neurodegenerative disease, methods of monitoring progression/regression of a neurodegenerative disease, methods of assessing efficacy of compositions for treating a neurodegenerative disease, methods of screening a composition for activity in modulating biomarkers associated with a neurodegenerative disease, methods of identifying potential drug targets for neurodegenerative diseases, and methods of treating a neurodegenerative disease. Such methods could be conducted as described herein with respect to Alzheimer's Disease.

Kits

A kit is envisaged for every method disclosed. The following description of a kit useful for diagnosing Alzheimer's disease in an individual by measuring the level of a biomarker in a biological sample therefore is not intended to be limiting and should not be construed that way.

The kit may comprise a negative control containing a biomarker at a concentration of about the concentration of the biomarker which is present in a biological sample of an individual who does not have Alzheimer's disease or does not have increased risk for Alzheimer's disease. The kit may also include a positive control containing the biomarker at a concentration of about the concentration of the biomarker which is present in a biological sample of an individual who as Alzheimer's disease or has increased risk for Alzheimer's disease.

Additionally, the kit includes a panel of biomarkers including one or more of the biomarkers shown in Table 2. However, the invention should not be limited to only these markers disclosed herein (e.g., Table 2) because a skilled artisan when armed with the present disclosure would be able identify additional markers that can be used as indicators for Alzheimer's disease. For example, as discussed in the Examples, a test sample and a control sample can be subjected to any commercially available panel comprising a plurality of markers and analyzed according to the statistic models disclosed herein to identify markers associated with AD.

Moreover, the disclosure presented here demonstrates a high degree of correlation amongst certain markers for identifying AD in an individual. For example, NT-proBNP showed high correlation with IGFBP2 and beta2 microglobulin suggesting these analytes may be linked in an AD specific pathophysiological pathway. Without wishing to be bound by any particular theory, it is believed that analytes that exhibit high correlation can be interchanged with other highly correlated analytes and still obtain similar differentiation performance. Thus, markers that are associated with an AD specific pathophysiological pathway can be interchangeable. Accordingly, correlation amongst the markers of the invention provides means to identify other related markers associated with the specific pathophysiological pathway as being indicators for

Alzheimer's disease.

One or more of the biomarkers in each or all of Tables 2, 3, 7, 8, 9, 10, 11, 12 or any fraction thereof can be included in the kit. In another aspect, other factors that predict for of AD can be included in the kit. Such factors include but are not limited to ApoE genotype (e2/e2, e2/e3, e2/e4, e3/e3, e3/e4 and e4/e4).

In one embodiment, the kit comprises a panel comprising at least Cortisol, Pancreatic Polypeptide, osteopontin, IGF BP2, and Resistin. In another embodiment, the kit comprises a panel comprising at least Resistin, e3/e3, Pancreatic Polypeptide, e3/e4, ApoD, G-CSF, MlP lbeta, ApoE, NrCAM, MMP2, Cortisol, PDGF-BB, e2/e3, MMP 1, 1-309.

In another embodiment, the kit comprises a panel comprising Alpha- 1 Microglobulin, Angiopoietin-2, Apolipoprotein E, Beta-2 Microglobulin, BLC, Cortisol, E-Selectin, FAS, Fatty Acid Binding Protein, Hepatocyte Growth Factor, IGF BP-2, Interleukin 10, NT-Pro-BNP, Osteopontin, Pancreatic Polypeptide, PAPP-A, Resistin, Stem Cell Factor, Tenascin C, Thrombomodulin, TIMP-1, VCAM-1, VEGF, and Von Willebrand Factor.

In another embodiment, the kit comprises reagents to assess age, gender,

ApoE genotype (e2/e2, e2/e3, e2/e4, e3/e3, e3/e4 and e4/e4) of the individual.

The kit of the invention can be used to assess the status of Alzheimer's disease in an individual, e.g., to diagnose Alzheimer's disease or to assess the degree of Alzheimer's disease in the individual. The phrase "Alzheimer's disease status" includes any distinguishable manifestation of the disease, including non-Alzheimer's disease, e.g., normal or non-demented. For example, disease status includes, without limitation, the presence or absence of Alzheimer's disease (e.g., Alzheimer's disease v. non-Alzheimer's disease), the risk of developing disease, the stage of the disease, the progress of disease (e.g., progress of disease or remission of disease over time) and the effectiveness or response to treatment of disease. Based on this status, further procedures may be indicated, including additional diagnostic tests or therapeutic procedures or regimens.

Furthermore, the kit includes an instructional material for use in the diagnosis of Alzheimer's disease in an individual. The instructional material can be a publication, a recording, a diagram, or any other medium of expression which can be used to communicate the usefulness of the method of the invention in the kit for assessment of Alzheimer's disease risk in a individual. The instructional material of the kit of the invention may, for example, be affixed to a container which contains other contents of the kit, or be shipped together with a container which contains the kit. Alternatively, the instructional material may be shipped separately from the container with the intention that the instructional material and the contents of the kit be used cooperatively by the recipient.

EXAMPLES The invention is further described in detail by reference to the following experimental examples. These examples are provided for purposes of illustration only, and are not intended to be limiting unless otherwise specified. Thus, the invention should in no way be construed as being limited to the following examples, but rather, should be construed to encompass any and all variations which become evident as a result of the teaching provided herein.

The experiments disclosed herein were designed to develop an assay to identify universal accepted biomarkers for diagnosing, monitoring and staging neurodegenerative diseases such as Alzheimer's disease that are fast, more accurate, and less expensive. The results presented herein demonstrate that a diagnostic assay has been developed that can detect among other things early onset of Alzheimer's disease.

Example 1 : Assessing diagnostic accuracy of a plasma based multiplexed immunoassay panel to identify Alzheimer's disease

Currently, prodromal Alzheimer's can be identified through the use of cognitive tests combined with cerebrospinal fluid (CSF) and imaging biomarkers.

However, such tests are often costly and invasive and are difficult to implement as screening tools to identify individual s suitable for clinical trial study. The identification of a screening blood test would significantly enable prevention studies in early

Alzheimer's disease. The present study examines the suitability of a multiplexed plasma panel as a screening tool to differentiate Alzheimer's disease from age-matched controls.

Briefly, a luminex based multiplex immunoassay panel consisting of 190 analytes was utilized in plasma samples derived from the University of Pennsylvania. The strategy to identify a plasma based panel included multivariate based approaches combined with examining changes in both CSF in both autopsy confirmed and non- autopsy confirmed individuals from the UPenn biobank. Final analytes were chosen based upon high performance in multivariate models and upon expression in CSF from autopsy confirmed individuals.

Multivariate training/test set models using all 190 analytes suggested Alzheimer's disease individuals could be differentiated from controls using analytes on the panel. Partial least square and shrunked centroid algorithms exhibited the best performance. Many of the top analytes identified in the multivariate and ANOVA analysis were also expressed in CSF obtained from autopsy confirmed individuals.

Multivariate and CSF results were combined to generate a final panel consisting of 24 multiplex analytes plus age, gender and ApoE genetic status (6 co-variates e2/e2; e2/e3; e2/e4; e3/e4; e3/e3; e4/e4). Performance of the panel in differentiating Alzheimer's disease from controls in the entire UPenn sample set using a logistic regression analysis exhibited 80% sensitivity 80% specificity.

Multivariate approaches combined with CSF-plasma expression and pathway analysis were utilized to identify a 32 variate algorithm that differentiates Alzheimer's disease from age-matched controls with high sensitivity suitable for preliminary screening of individuals who might be eligible for clinical trial enrollment in early Alzheimer's disease studies.

The materials and methods employed in these experiments are now described.

University of Pennsylvania Sample Set

Individuals were recruited and longitudinally followed at the University of Pennsylvania (UPenn). All protocols were approved by the University of Pennsylvania Review Board and all individuals, and in cases of dementia individuals, caregivers, provided an informed consent for utilization of biofluids for exploratory research purposes. Clinical information was de-identified in compliance with privacy regulations. Individuals diagnosed with clinical Alzheimer's disease and frontotemporal lobar degeneration (FTLD), progressive supranuclear palsy (PSP) and corticobasal syndrome (CBS) were longitudinally followed in the ALS center and Frontotemporal dementia clinic. Each individual underwent detailed neurological and laboratory examination to ensure the accuracy of clinical diagnosis according to established criteria for Alzheimer's disease and FTLD. Plasma samples were obtained from 92 Alzheimer's disease individuals, 126 normal healthy controls and 48 with other forms of dementia. Autopsy confirmed CSF samples were obtained from 70 Alzheimer's disease individuals and 25 individuals with other dementia (TDP, Tau and DLB). 32 CSF samples from cognitively normal, non-autopsied elderly controls were also tested. ApoE genotyping

ApoE genotyping was done for all individuals using EDTA blood samples. Taqman quantitative PCR assays were used for genotyping APOE nucleotides 334 T/C and 472 C/T with an ABI 7900 real time thermocylcer using DNA freshly prepared from EDTA whole blood. Luminex xMAP panel

Plasma samples were analyzed using a commercially available multiplexed luminex human discovery xMAP panel from Rules Based Medicine (RBM, Austin, Texas). All assays were validated according to CLIA standards. In brief, the luminex technology multiplexes immunoassays on the surface of polystyrene microsphere beads. The microsphere beads are loaded with a ratio of two spectrally distinct flourochromes yielding up to 100 uniquely color-coded beads. The beads are coated with capture antibodies specific for the assay and run in either standard sandwich or competitive immunoassay format. Capture-antibody microspheres are incubated with blocking solution and diluted plasma sample or calibration controls for one hour. Beads are rinsed and biotinylated detection reagent added. Streptavidin-phycoerthyrin is then added to each well and incubated for 60 minutes. Following wash steps, beads are resuspended in reading solution and read on the luminex instrument. Details of each assay procedure are provided by RBM in validation reports that are compliant with CLIA standards. Some of the assays have defined a lower limit of quantitation. For the purposes of the present study, the lower limit of detection was utilized. Validation details of all the assays are available from Rules Based Medicine (http://www.rulesbasedmediciiie.com). Plasma Collection Methods

In brief, whole blood was collected into 10ml BD lavender top K₂EDTA(18mg spray coated or 15% solution) coated vacutainers. Samples were gently mixed by inversion and spun at 3000 rmp for 15 minutes at room temperature within one hour of collection. Immediately after centrifugation, plasma was transferred to a polypropylene transfer tube, placed in dry ice and allowed to completely freeze. Samples were then stored at -80C° until analysis. A 0.5ml aliquot that had not been allowed any freeze-thaw cycles was shipped to RBM for analysis.

Statistical Methods

For purposes of the univariate analysis, analytes that had more than 10% missing values were excluded from the analysis. Values that were reported as LOW (below the lowest assay limit which is defined by RBM as a value below the lowest calibrator for the individual assay) were imputed by taking the reported low detection limit for individual assays (LDD) and dividing by 2. Individual analyte distributions were tested for normalcy using Box-Cox analysis. If necessary, analytes were log transformed and checked again for normalcy. The majority of analytes exhibited non-normal distributions and were log transformed. For purposes of the univariate analysis, multidimensional scaling and Mahalanobis distances were used to detect outliers. Outliers were defined as 5 standard deviations beyond the mean and were replaced with the 5 STD value. A major concern in the analysis was control of type I error rate due to relatively large number of plasma proteins in the multiplex panel. False discovery (FDR) corrections were applied to p-values to account for the multiple comparisons. ANOVA models including diagnosis, age, gender and ApoE4 allele status were utilized in the final model. ApoE allele status was classified into 6 subcategories; e2/e2, e2/e3, e2/e4, e3/e3, e3/e4, e4/e4.

Multiple marker analysis approach was used to build relationships between disease groups. Approaches applied included logistic regression, boosted tree models, FDA, KNN, Naive Bayes, Partial Least Squares, Random Forests, shrunken centroids, sparse partial least squares and support vector machine approaches. Three models were attempted using different co-variates: 1) age, gender and ApoE4 genetic status alone, 2) RBM assay results only and 3) RBM assay results plus age, gender and ApoE4 allele status in order to understand the predictive ability of the assays beyond demographic and genotype information currently used to assess risk. Prior to modeling, pre-filtering of markers using K resampling iterations in an unsupervised fashion were performed on UPenn dataset. This included splitting UPenn data into training and test sets and then applying an unsupervised filter on the predictors. The predictive model using only UPenn data was further built and tuned on the training set. The model was then tested on the UPenn test dataset as an independent dataset separate from the trained dataset.

Classification accuracy and kappa features of the performance were saved. All feature selection routines were extensively cross validated using methods described by Ambroise and McLachlan 2002 PNAS). Measures of marker importance were biased towards those that use uncertainty (logistic regression slope tests) as opposed to those that did not (random forest variable importance etc).

The results of the experiments are now described.

Demographics of the UPenn cohort

Table 1 summarizes demographic characteristics of the UPenn sample set. Two hundred and sixty six plasma samples were shipped for multiplex analysis. Of the 266, 92 were from Alzheimer's disease, 126 were from healthy elderly controls and 48 were from individuals with other forms of dementia. Control and dementia individuals from the UPenn cohort were significantly younger than Alzheimer's disease. Thus age was included as a co-variate in all models. Amongst the groups, females were more highly represented, but there were no significant differences in gender ratios between the groups. Gender was also included as a co-variate in all the models. The prevalence of ApoE4 allele in the Alzheimer's disease group was 63%, very similar to what was observed in general Alzheimer's disease population. In addition, 102 CSF samples from autopsy confirmed individuals and 32 CSF samples from non-autopsy confirmed CSF controls were included in the analysis to determine whether changes in plasma analytes were reflected by changes in CSF from autopsy confirmed individuals. Age, gender and ApoE prevalence did not differ between the plasma and CSF cohorts.

Table 1 : Demographics University of Pennsylvania Sample Set

*p<0.05 when compared to Alzheimer's disease

Table 2 summarizes all the analytes included in the RBM panel. Light gray highlights those analytes that had more than 10% missing or had more than 10% LOW values. Multivariate approaches were combined with CSF expression patterns for feature selection. The top 24 performers are highlighted in dark gray (Table 2). The multivariate analyses in this context are the predictive models that were performed (e.g. Random forest, PLC, etc.) for the goal of predicting the outcome.

Table 2: Analyte and final model summary. Analytes with more than 10% missing or had more than 10% LOW values (highlighted in light gray). Final panel (highlighted in dark grey) were selected based upon multivariate results and upon expression from auto sy confirmed individuals.

A number of multivariate approaches were applied to aid in final feature selection. The strategy for model generation relied upon training with 75% of data and then testing on 25% of the remaining cohort data. Multivariate strategies included a shrunken centroid algorithm developed for predictive analysis of microarrays (PAM) (Tusher et al, 2001 Proc Natl Acad Sci U S A. 98(9):5116-5121), FDA, KNN, logistic regression, Naive Bayes, partial least squares, random forest, sparse partial least square, boosted trees, linear discriminant analysis (data not shown), principal components (data not shown) and support vector machine (Vapnik et al, 1999 IEEE Trans Neural Netw. 10(5):988-999). ROC curves from the analysis are shown in Figure 1. Partial least squares and shrunken centroid approaches appeared to give maximal performance in the training and test sets. Because the cohorts were biased in terms of enrollment, all models included age, gender and ApoE4 allele status as co-variates. Figure 1 lists the top 30 contributors to the partial least square model. Genotype was an important contributor and appeared in the top list. As a result, genotype models were used as baseline models for subsequent comparisons.

ANOVA analysis were then conducted on all the analytes to determine whether changes were statistically significant between Alzheimer's disease and age- matched controls and Alzheimer's disease and other dementias. Although many analytes were statistically different in the Alzheimer's disease group compared to either control or other dementia groups (Table 3), there were few analytes that exhibited robust differences (e.g. more than 1.5X) compared to the Alzheimer's disease group. Interestingly, a subset of analytes appeared to differentiate Alzheimer's disease from other types of dementias including a number of the apolipoproteins and inflammatory factors.

Table 3 : Summary of analyte levels in UPenn Plasma and ANOVA Results: Age and gender included as co-variates and p values reported for Alzheimer's disease v control

0 8.52 + 2.60 0 7.23 1.51

Π- Ι 2ρ40 Alzheimer' s disease 0.0054 Thrombospon Alzheimer's disease 0.0491 <ng ml) 0.53 + 0.25 din- 1 (ng.ml) 293.2 + 11228.5 <0.0001*

C 0.44 + 0.23 C 7581.5 + 9044.7

0 0.46 + 0.23 0 14217.2 + 12022.1

II .- 1 (pg'ml) Alzheimer's disease 0.0448 VK(il Alzheimer's disease 0.0956

101.16 + 41.82 0.0013* (pg'ml) 605.57 + 316.20

C 88.69 + 35.28 C 531.61 + 267.38

0 78.98 + 34.87 0 607.14 + 248.07

II - I ra (pg'ml) Alzheimer's disease 0.7702 Von Alzheimer's disease 0.3465

85.64 + 66.19 0.0010* Willelirand 33.04 + 18.23 0.0193* C 82.15 + 51.83 1 -^'actor C 28.13 + 13.68 0 121.88 + 90.69 (ug ml) 0 36.24 + 16.29

A comparison analysis was then conducted to examine expression in CSF from autopsy confirmed individuals with analytes that exhibited statistical significance by ANOVA analysis or were identified as top differentiators in the multivariate models. Figure 2 summarizes expression of a subset.

A logistic regression model was utilized to compare performance of a baseline model (age, gender and ApoE allele status) to a model including only the 24 top plasma analytes and a model including all 24 plasma analytes plus the baseline model. All models were benchmarked at 80% sensitivity to allow comparisons across the models. In general, the baseline demographic model using only age, gender and ApoE4 allele showed positive predictive value of 52%. A model including all 24 analytes showed a positive predictive value of 64% and a model including all 24 analytes and baseline demographics showed a reasonable positive predictive value of 75% (see Table 4). ROC curves for each of the models are depicted in Figure 3. The final logistic regression model including age, gender, ApoE allele and 24 plasma analytes yielded a 80% sensitivity and 80% specificity in differentiating Alzheimer's disease from controls. Model showed very good performance in differentiating Alzheimer's disease from other forms of dementia as well. A model was attempted using 12 of the 18 analytes described by Ray et al. (2007 Nat Med. 13(1 1): 1359-1362) (Angiopoietin-2, EGF, G-CSF, ICAM- 1, Interleukin 3, Interleukin 8/CXCL8, M-CSF, PDGF-BB, PARC/CCL18,

RANTES/CCL5 and TNF-alpha). Two analytes (Interleukin la, Interleukin- 11) were below the limit of detection and not included in the analysis. Four analytes

(MIP5/CCL15, GDNF, IGF-BP6, TRAIL-R4), were not present on the panel, but related analytes including IGF-BP2 and TRAIL R3 were represented. Internal studies using the Ray et al. (2007 Nat Med. 13(1 1): 1359-1362), data and excluding MIP5, GDNF, IGF- BP6 and TRAIL-R4 did not degrade the performance of the model (data not shown). Preliminary results could not confirm performance reported in Ray et al. (2007 Nat Med. 13(1 1): 1359-1362), using multiplex analytes alone (data not shown). However, inclusion of demographic information improved performance close to levels of the 32 co-variate model described in Figure 3.

Est mates enc mar e aga nst sens t v ty n a mo e s

Correlation matrix analysis in the Alzheimer's disease samples showed a high degree of correlation amongst the 24 analytes (Figure 4A). Figure 4B is a chart also demonstrating high correlation of the analytes with each other and that there is high correlation with other analytes on the panel. Without wishing to be bound by any particular theory, it is believed that any of the individual 24 analyte can be switched out with any analyte that showed significant correlation with another analyte on the panel (p<0.05). For example, NT-proBNP showed high correlation with IGFBP2 and beta2 microglobulin suggesting these analytes may be linked in an Alzheimer's disease specific pathophysiological pathway. Many of the analytes in the 190 panel showed high correlation and it was possible to interchange the top 24 analytes with other highly correlated analytes and obtain similar differentiation performance (data not shown).

A subanalysis was completed to examine the contribution of ApoE genotype to multiplex analyte expression. There has been some controversy in the literature around whether individuals with an ApoE4 allele exhibit altered protein levels. In the current dataset, ApoE plasma protein levels differed based upon ApoE allele genotype irrespective of diagnostic status. For example, ApoE4/E4 homozygotes showed low levels of plasma ApoE protein compared to ApoE2/E2 homozygotes (Figure 5). The phenomena was not limited to ApoE protein levels as a similar pattern was noted in plasma C Reactive Protein levels (Figure 3). IL-15 levels were elevated in ApoE4/E4 homozygotes. These data suggest an ApoE endophenotype is associated with specific patterns of expression of apolipoprotein and cytokine analytes irrespective of diagnosis.

Figure 5 compares apolipoprotein E levels in individuals as a function of their ApoE genotype. In both datasets, individuals with either an e3/e4 or e4/e4 genotype showed lower apolipoprotein levels compared to individuals with e2/e2, e3/e2 or e3/e3 genotypes. Individuals with an e2/e4 genotype exhibited levels more closely matched to e3 and e2 groups. Although the numbers are very small, individuals homozygous for E2 exhibited much higher protein levels. Alzheimer's disease individuals were present in each of the ApoE genetic groups with the exception of e2 homozygotes. These data suggest that apolipoprotein levels do differ dependent upon ApoE genotype and that levels are lowest in individuals with an e3/e4 or and e4/e4 genotype.

Plasma Biomarkers for Diagnosis of Alzheimer's disease

Identification of Alzheimer's disease individuals prior to the onset of dementia will be critical for successful development of Alzheimer's disease disease modifying drug. Although there are CSF and imaging research tools currently available to identify individuals at risk of progressing to dementia, these tools are difficult to implement as screening tools for individual enrollment due to the cost and invasive nature of CSF testing. A simple blood test would aid in identifying those eligible for more confirmatory biomarker and cognitive testing and would help raise awareness around dementia prevention.

The current study examined the utility of a commercially available multiplex blood panel in differentiating Alzheimer's disease from controls using samples obtained from the University of Pennsylvania. Initial studies were initiated based upon a report by Ray et al, (2007 Nat Med. 13(11): 1359-1362) that an 18 analyte panel could differentiate Alzheimer's disease from controls. Performance results of the a subset of the Ray et al. (2007 Nat Med. 13(1 1): 1359-1362), analytes in the UPenn sample set could only be reproduced if age and ApoE genetic allele status were included in the model suggesting performance was largely driven by age and ApoE status.

Multiple approaches were used for feature selection from the 190 analyte.

The top lists of obtained from multivariate approaches were compared to ANOVA lists and lists obtained from examining expression of analytes in CSF from autopsy confirmed individuals. A final model containing 24 analytes was selected for further analysis. Inclusion of the 24 multiplex analytes improved performance over baseline demographic models in identifying individuals with Alzheimer's disease. Logistic regression estimates of model performance showed 80% sensitivity and 80% specificity in differentiating Alzheimer's disease from controls. However, caution should be exercised as these estimates are likely to degrade in a population setting and when tested in independent datasets. Correlation analysis utilized in the current study to identify analyte redundancy in the panel was actually quite revealing as many of the analytes showed strong correlations. The finding may be indicative of a biological signaling cascade activated during pathological sequeale associated with dementia. Further pathway analysis may shed insight into the underlying pathophysiological and specific endophenotypes associated with Alzheimer's disease and other forms dementia.

Finally, the current study demonstrated that plasma apolipoprotein E levels are altered dependent upon genotypic status. Specifically, individuals with an e3/e4 or e4/e4 genotype exhibited lower apolipoprotein E and C reactive protein and higher IL-14 levels compared to other ApoE genotypes. Reports of altered ApoE protein levels in Alzheimer's disease individuals have been controversial and studies examining changes in message levels often contradict protein findings (Kim et al, 2009 Neuron 63(3):287- 303). Apolipoprotein E is a glycoprotein involved in lipid transport and a ligand for receptor mediated endocytosis. It is one of the major high density lipid (HDL) components in brain and believed to be critical for cholesterol transport involved in brain synaptic turnover. ApoE is the only gene consistently shown to increase the risk of late onset Alzheimer's disease and E4 is believed to confer greater risk while E2 is believed to be protective for dementia. The finding that ApoE4 levels are lower in plasma in Alzheimer's disease individuals may have relevance to disease if levels are altered in a similar way in CNS. Low levels may impact normal synaptic function and membrane recycling at the synapse (Hirsch-Reinshagen et al., 2009 Mol Cell Biochem. 326(1- 2): 121-129). Interestingly, a novel genetic risk factor has been identified in another apolipoprotein, ApoJ also known as clusterin. The current study did not identify any changes in ApoJ levels in either cohort and additional studies in CSF are needed to further understand relevance.

The current findings support that a plasma based signature for Alzheimer's disease can be detected in plasma and that signature can differentiate Alzheimer's disease from healthy controls and other forms of dementia. The signature can improve diagnostic accuracy over age and ApoE4 allele status alone, but does not rise to the level of a standalone diagnostic tool. The panel may, when combined with a simple memory test, have utility as a screening tool for individuals with a cognitive complaint who would then be eligible for more confirmatory CSF and imaging testing.

Numerous proteomic approaches have long sought a diagnostic signature in plasma or serum (Ray et al, 2007 Nat Med. 13(1 1): 1359-1362, Hye et al, 2006 Brain 129(Pt l l):3042-3050; Irizarry et al, 2004 NeuroRx. l(2):226-234; Song et al, 2009 Brain Res Rev.; Zhang et al, Proteomics 4(l):244-256). Unfortunately, much of the current proteomic technological approaches have been qualitative in nature and more quantitative confirmation in independent studies using more confirmative immunoassay approaches have been elusive. Here, a quantitative approach was utilized to assess performance of a multiplex panel in differentiating Alzheimer's disease from controls. The panel showed a significant, improvement over use of age, gender and ApoE4 allele status alone.

Example 2: RBM Biomarker Analysis

The data set consisted of 165 characteristics (i.e. variables) collected on

218 individuals. Of these individuals, 92 were determined to be Alzheimer's disease and

126 were controls.

The following variables were removed due to missing values or other factors: IgE, IL12p70, IL17E, IL1 alpha, Prostate Specfic Ag, Tissue Factor. After filtering, there were 164 variables used for prediction. The data were split into a training set (n=175) and a test set(n=43) in a manner that preserves the frequency distribution between Alzheimer's disease and control individuals. Where needed, the training and test sets were centered and scaled. Models

A variety of statistical models were used to predict the data:

• Random Forests: A sequence of 3 values for the number of retained variables were used to tune the model (ranging from 2 to 2). Breiman. Random forests. Machine learning (2001) vol. 45 (l) pp. 5-32.

· Boosted Tree: candidate models had interaction depths raining from 2 to 10, while the number of boosting iterations ranged from 100 to 2000. Friedman. Greedy function approximation: a gradient boosting machine. Annals of Statistics (2001) pp. 1 189-1232.

• Partial Least Squares: the number of PLS components varied from 1 to 20.

• Sparse Partial Least Squares: the number of PLS components varied from 1 to 20 and the regularization parameter ranged from 0.1 to 0.9. Partial Least Squares: Martens and

Naeses, Multivariate calibration. (1989) Wiley; Sparse Partial Least Squares: Chun H, Keles S: Sparse partial least squares for simultaneous dimension reduction and variable selection. Journal of the Royal Statistical Society 2009, 182(l):79-90 • Flexible Discriminant Analysis: candidate models for the MARS basis function used from 2 to 63 terms. Flexible Discriminant Analysis: Hastie et al. Flexible Discriminant Analysis by Optimal Scoring. Journal of the American statistical association (1994) vol. 89 (428).

· Naive Bayes: models were computed using a Gaussian kernel or a non-parametric density estimate.

• Nearest Shrunken Centroids: candidate models used threshold parameters ranging from 0.092 to 2.583. Nearest Shrunken Centroids: Tibshirani et al. Class prediction by nearest shrunken centroids, with applications to DNA microarrays. Statistical Science (2003) pp. 104-117.

• Logistic Regression: weight decay/regularization was used to stabilize the model.

Parameters ranging from 0 to 0.1 were tested. The Elements of Statistical Learning: Data Mining, Inference, and Prediction (Second Edition) by Trevor Hastie, Robert Tibshirani and Jerome Friedman (2009).

· K-Nearest Neighbors: the number of neighbors tested ranged from 5 to 23. The

Elements of Statistical Learning: Data Mining, Inference, and Prediction (Second Edition) by Trevor Hastie, Robert Tibshirani and Jerome Friedman (2009).

• Support Vector Machines: a radial basis function was used with cost parameter values ranging from 0.1 to 100. The Elements of Statistical Learning: Data Mining, Inference, and Prediction (Second Edition) by Trevor Hastie, Robert Tibshirani and Jerome

Friedman (2009).

For each model, the tuning parameter used 50 iterations of the bootstrap to compute resampled estimates of the sensitivity and specificity. Previous modeling for these data showed that high specificity can be easily attained, so the final model was chosen by maximizing the sensitivity value across the candidate models.

Results

Sensitivity and specificity were used to evaluate the data on the training and test sets. On the training set, 50 iterations of the bootstrap were used to get reasonable estimates of performance and to differentiate models. Sensitivity, specificity and the area under the ROC curve were also calculated on the 43 samples in the test set. Table 5 shows the results for the training data across all the models. Table 6 also contains the test set results. Figure 6 shows these results in graphical terms. Table 5: Training set resampling results across various models

:··.:)¾ S;|5«

Boosted I se O.MS e '.7 1 4

FDA 0.537 aS O

K N

logistic ί'ΛνΚ

FLfS 0,201

Jt-iiid_' iiii K Ksis 0,76y

Slirnuko GeiilMid-i ') 4 1

Spars PUS itMt) « '.7¾>

SVM (L¾63 ϊ . 67

Table 6: Test set results across various models

mo el A UG

O.ME) 0.80 0.73!

FDA 0.611 o,s¾ 0.764

Logistic Ri.-g 0.611 0,72 0.71»

Naivsj B& is 0.611 I i t

FLS «1667 o <:o

i :. -n l. -ni I¾reste *> ;·.!::· 0.72 ■ ¹ il

tn in: l. fi ·^' »r I . > 0'. 0.723

:· i: Ρ-> · FLB 0.611 0,72 0.7-19

0.667 0,84 0.7iB

Figure 1 shows the ROC curves for the test set with each model. The solid black point shows the sensitivity/specificity combination that corresponds to using a cutoff of 0.50 on the probability estimate of being Alzheimer's disease. The solid square indicates an alternate choice based on getting the sensitivity as close as possible. The test in each panel shows the alternative cutoff and the corresponding sensitivity/specificity estimates.

Without wishing to be bound by any particular theory, it is believed that caution should be taken when interpreting the results from the alternate cutoff. Since ROC curve analysis is part of the modeling process, any new cutoff should be evaluated on a separate data set.

Figure 7 also shows the sensitivity and specificity profiles across various probability cutoffs. This plot can demonstrate how adaptable each model is to alternative cutoffs. It was observed from the models that were run on these data, the partial least squares (PLS) model appeared to have the best combination of sensitivity and specificity, as well as a high test set ROC AUC value. The PLS model used all the predictors in the data set.

Table 7 has a list of the top 15 predictors that contribute to the model. The overall distribution of the scores is shown in Figure 8. In Figure 8, there is not clear cluster of variables with "high" and "low" importance. The rankings of the genotype variables were not in the top 15.

Table 7: The top 15 most important predictors for the PLS model (from most important to least)

Overall

ΓΊ:ν·

(m

-■U-i i Hi.'

ApoD ^! IIS

CLCSF if.011;

;p :.b-n., c \'<<;

A; >,L !i.0100

··:,·<: Λ VI if.OlOO

UMP2 Oijij!!iiO

FDGP_BB 'i,0OE. i

S^'.U^'jiiJO

iM^'P!

t_3(¾

For the top 15, Figure 9 shows the correlation patterns between the variables and Figure 8 shows the distribution of the training set data for each predictor. Figure 10 shows no single predictor that significantly differentiates Alzheimer's disease from control. This also supports the idea that the models require a large number of predictors to achieve adequate performance.

Table 8 and 9 is based on the univariate Analysis of Covariance

(ANCOVA) models comparing the two phenotypes (AD vs other) while adjusting also for age and gender of the subjects. The p-values supplied are based on the type III sums of squares, thereby reflecting the ability of the marker (on its own, not in conjunction with other markers) to distinguish between the two groups above and beyond the ability of age and gender to discriminate. Table 8: UPenn Plasma Univariate: Comparison Alzheimer's disease vs Control only

Response Log Alpha 1 -Antichymotrypsin Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.09757082 ₄.9286 0.0275

Age 1 1 0.00632811 0.3197 0.572₄

Gender 1 1 0.02083983 1.0527 0.3060

Response Log Alpha-1 Antitrypsin Effect Tests

Source N arm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.1165₄156 12.7513 0.000₄

Age 1 1 0.002966₄1 0.32₄6 0.5695

Gender 1 1 0.01₄36133 1.5713 0.2114

Response Log ACE Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.00893135 0.3298 0.5664

Age 1 1 0.06297906 2.3254 0.1288

Gender 1 1 0.13529560 4.9955 0.0264

Response Log Adiponectin Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.10072464 1.4883 0.2238

Age 1 1 0.20830292 3.0778 0.0808

Gender 1 1 0.95091021 14.0503 0.0002

Response Log Agouti Related Protein Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.05191313 0.6177 0.4328

Age 1 1 0.22554090 2.6838 0.1028

Gender 1 1 0.14469766 1.7218 0.1909

Response Log Alpha-2 Macroglobulin Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.01371995 2.6884 0.1025

Age 1 1 0.00418567 0.8202 0.3661

Gender 1 1 0.10168754 19.9256 <.0001

Response Log Alpha-Fetoprotein Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.0039838 0.0323 0.8575

Age 1 1 1.6238383 13.1662 0.0004

Gender 1 1 0.8761198 7.1036 0.0083

Response Log Alphal Microglobulin Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.03956080 2.3233 0.1289

Age 1 1 0.19675816 11.5548 0.0008

Gender 1 1 0.02769081 1.6262 0.2036

Response Log ANG-2 Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.00742880 0.2975 0.5860

Age 1 1 0.99784496 39.9622 <.0001

Gender 1 1 0.07306225 2.9260 0.0886

Response Log Angiotensinogen Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.0691295 0.0879 0.7672

Age 1 1 1.2130149 1.5418 0.2157

Gender 1 1 1.5113540 1.9210 0.1672

Response Log Apo A1 Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.13548238 5.7580 0.0173

Age 1 1 0.00010260 0.0044 0.9474

Gender 1 1 0.46759744 19.8730 <.0001

Response Log Apo A2 Effect Tests

Source Nparm Sum of Squares F Ratio Prob > F

DIAG 1 0.00150715 0.1137 0.7363

Age 1 0.02247936 1.6956 0.1943 Source N arm DF Sum of Squares F Ratio Prob > F Gender 1 1 0.0998₄367 7.5313 0.0066

Response Log ApoCI Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.00001020 0.0008

Age 1 1 0.0272₄053 2.0179

Gender 1 1 0.₄2₄93871 314785

Response Log Apo Clll Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.00055256 0.02₄₄

Age 1 1 0.00168₄50 0.07₄₄

Gender 1 1 0.27₄6₄797 12.1233

Response Log ApoH Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.10012685 8.1323

Age 1 1 0.00380708 0.3092

Gender 1 1 0.00001769 0.001₄

Response Log Apo A-IV Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.00032618 0.0078 0.9297 Age 1 1 0.06832₄97 1.6356 0.2023 Gender 1 1 0.08021122 1.9201 0.1673 Response Log ApoB Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.02689237 2.2022

Age 1 1 0.08758695 7.1725

Gender 1 1 0.0₄2885₄6 3.5119

Response Log ApoD Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.100019₄₄ 5.630₄

Age 1 1 0.00897956 0.5055

Gender 1 1 0.22521563 12.6780

Response Log ApoE Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.20639₄89 6.5366 0.0113

Age 1 1 0.02060278 0.6525 0.4201

Gender 1 1 0.63626015 20.1505 <.0001

Response Log ApoJ Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.0021₄321 0.₄₄80 0.5040

Age 1 1 0.01316376 2.7515 0.0986

Gender 1 1 0.08331881 17.4154 <.0001

Response Log AXL Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.00470212 0.2133 0.6446

Age 1 1 0.00018707 0.0085 0.9267

Gender 1 1 0.03260202 1.4791 0.2252

Response Log Beta2 Microglobulin Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.04007431 2.9127 0.0893 Age 1 1 0.66365011 48.2355 <.0001 Gender 1 1 0.00331402 0.2409 0.6241

Response Log BLC Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.04064608 0.5990

Age 1 1 0.65398649 9.6370

Gender 1 1 0.03176943 0.4681

Response Log BMP -6 Effect Tests

Source Nparm Sum of Squares F Ratio Prob > F DIAG 1 0.01049858 0.1622 0.6875 Source N arm DF Sum of Squares F Ratio Prob > F

Age 1 1 0.3009081_{4 4}.6502 0.0322

Gender 1 1 0.2678965₄ 4.1401 0.0431

Response Log BDNF Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.9739714 3.9635

Age 1 1 2.3079839 9.3922

Gender 1 1 0.0059720 0.0243

Response Log Betacellulin Effect Tests

Source N arm DF Sum of Squares F Ratio

DIAG 1 1 0.27314883 3.1594

Age 1 1 0.04122411 0.4768

Gender 1 1 0.04713463 0.5452

Response Clean C3 Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.03250361 1.5681

Age 1 1 0.05851025 2.8228

Gender 1 1 0.19054226 9.1927

Response Log Cancer Antigen 19-9 Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.0104371 0.0217 0.8830 Age 1 1 1.0640636 2.2148 0.1382 Gender 1 1 0.0913984 0.1902 0.6632

Response Log CD40 Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.11245269 5.5735 0.0191 Age 1 1 0.39083707 19.3712 <.0001 Gender 1 1 0.00015022 0.0074 0.9313

Response Log CD40 Ligand Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.3263141 3.4310

Age 1 1 1.2308276 12.9415

Gender 1 1 0.0092553 0.0973

Response Log CDL5 Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.06907743 2.6644 0.1041

Age 1 1 0.33110385 12.7710 0.0004

Gender 1 1 0.00551303 0.2126 0.6452

Response Log Carcinoembryonic Antigen Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.0268749 0.3593 0.5495 Age 1 1 1.3395152 17.9082 <.0001 Gender 1 1 0.0015312 0.0205 0.8864

Response Log CgA Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.0153585 0.2311

Age 1 1 2.4438349 36.7693

Gender 1 1 0.0002601 0.0039

Response Log CKMB Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.29278936 4.5918

Age 1 1 0.51286060 8.0431

Gender 1 1 0.70203279 11.0099

Response Log CNTF Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.09135214 1.1767

Age 1 1 0.09781011 1.2599

Gender 1 1 0.09022008 1.1621

Response Log Complement Factor H Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F Source N arm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.00₄7₄600 0.150₄ 0.6985 Age 1 1 0.0₄2₄1890 1.3₄₄5 0.2₄75 Gender 1 1 0.00213276 0.0676 0.7951

Response Log CTGF Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.0829332₄ 2.16₄6

Age 1 1 0.007₄1676 0.1936

Gender 1 1 0.23071753 6.0218

Response Log Cortisol Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.22555210 13.3566

Age 1 1 0.00₄30777 0.2551

Gender 1 1 0.10357875 6.1337

Response Log C-Peptide Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.00075280 0.0078

Age 1 1 0.81716606 84666

Gender 1 1 0.6206₄905 64305

Response Log CRP Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 1.4184758 4.6475

Age 1 1 0.2229228 0.7304

Gender 1 1 0.9091701 2.9788

Response Log Cystatin Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.06347550 4.2353

Age 1 1 0.32585564 21.7424

Gender 1 1 0.03909462 2.6086

Response Log EGF Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.6643216 5.4834

Age 1 1 1.4071244 11.6147

Gender 1 1 0.0393685 0.3250

Response Log EGF-R Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.04021177 1.8645 0.1735

Age 1 1 0.20639230 9.5699 0.0022

Gender 1 1 0.00653961 0.3032 0.5824

Response Log ENA-78 Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.28983488 1.9380 0.1653

Age 1 1 0.26394974 1.7649 0.1854

Gender 1 1 0.33790605 2.2594 0.1343

Response Log Endothelin-1 Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.00518124 0.0836

Age 1 1 0.01042439 0.1682

Gender 1 1 0.09779083 1.5782

Response Log EN-RAGE Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.48257884 6.4912

Age 1 1 0.12175491 1.6377

Gender 1 1 0.16606285 2.2337

Response Log Eotaxin Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.00089570 0.0099

Age 1 1 0.13198730 1.4561

Gender 1 1 0.04421998 0.4878 Response Log E-Selectin Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.1₄921791 4.1136

Age 1 1 0.00233491 0.0644

Gender 1 1 0.01617998 0.4461

Response Log FABP Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.0136922 0.0673

Age 1 1 6.2144717 30.5362

Gender 1 1 1.2982004 6.3790

Response Clean Factor VII Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 7896.37 0.2188 0.6404

Age 1 1 3266.43 0.0905 0.7638

Gender 1 1 606208.80 16.7959 <.0001

Response Log FAS Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.05636966 3.8801

Age 1 1 0.15077612 10.3784

Gender 1 1 0.10987253 7.5629

Response Log Fas-Ligand Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.02145776 0.4647 0.4962 Age 1 1 0.22218847 4.8122 0.0293 Gender 1 1 0.15718221 3.4043 0.0664

Response Log Ferritin Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.0401446 0.2679 0.6053 Age 1 1 0.5584810 3.7264 0.0549 Gender 1 1 1.4988784 10.0012 0.0018

Response Log Fetuin A Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.00513336 0.8342

Age 1 1 0.07751276 12.5969

Gender 1 1 0.02472305 4.0179

Response Log Fibrinogen Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.06487349 4.1894

Age 1 1 0.00689806 0.4455

Gender 1 1 0.06763528 4.3678

Response Log FSH Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.0022382 0.0650 0.7991

Age 1 1 0.2907121 8.4361 0.0041

Gender 1 1 6.3655922 184.7220 <.0001

Response Log G-CSF Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.33721686 8.3831 0.0042

Age 1 1 0.04529264 1.1260 0.2898

Gender 1 1 0.27247284 6.7736 0.0099

Response Log Growth Hormone Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.1319637 0.3275 0.5677

Age 1 1 4.3569639 10.8132 0.0012

Gender 1 1 1.8833330 4.6741 0.0317

Response Log GLP-1 Total Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.00196054 0.0675 0.7953

Age 1 1 0.00007370 0.0025 0.9599

Gender 1 1 0.07023159 2.4168 0.1215 Response Log Glucagon Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.06088236 1.5376

Age 1 1 0.05101071 1.2883

Gender 1 1 0.119₄₄776 3.0166

Response Log GST alpha Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.51211036 3.7203

Age 1 1 0.30192565 2.193₄

Gender 1 1 0.515₄0279 3.7₄₄2

Response Log GRO-alpha Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.39391010 ₄.0260

Age 1 1 0.0506126₄ 0.5173

Gender 1 1 0.00258595 0.026₄

Response Log Haptoglobulin Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.20300₄86 0.7516

Age 1 1 0.0₄₄623₄₄ 0.1652

Gender 1 1 0.58762986 2.1757

Response Log HB-EGF Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.1₄5₄7₄3 0.7115 0.3999

Age 1 1 1.1304977 5.5290 0.0196

Gender 1 1 0.4959013 2.4253 0.1209

Response Log HCC-4 Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.03513513 0.7691 0.3815

Age 1 1 0.12077438 2.6436 0.1054

Gender 1 1 0.01826546 0.3998 0.5279

Response Log HGF Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.00449682 0.2871

Age 1 1 0.03281238 2.0953

Gender 1 1 0.06866183 4.3845

Response Log HSP60 Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.21681869 2.5026

Age 1 1 0.02382186 0.2750

Gender 1 1 0.53678302 6.1957

Response Log I-309 Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 1.0215429 4.4009

Age 1 1 0.1730087 0.7453

Gender 1 1 0.0013619 0.0059

Response Log ICAM-1 Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.00169748 0.0885 0.7664

Age 1 1 0.04991320 2.6011 0.1083

Gender 1 1 0.08234349 4.2912 0.0395

Response Log IgA Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.24489265 5.9353 0.0157

Age 1 1 0.01852248 0.4489 0.5036

Gender 1 1 0.00208025 0.0504 0.8226

Response Log IgE Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.3139331 0.6930 0.4061

Age 1 1 0.7377087 1.6285 0.2033

Gender 1 1 2.5200239 5.5630 0.0193 Response Log IGFBP-2 Effect Tests

Source N arm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.6₄20886 10.7303 0.0012 Age 1 1 2.2682₄07 37.9059 <.0001 Gender 1 1 0.0006157 0.0103 0.9193

Response Log IgM Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.00878₄21 0.128₄

Age 1 1 0.08182886 1.1958

Gender 1 1 0.07₄0₄381 1.0820

Response Log IL-10 Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.10788₄28 3.6₄0₄

Age 1 1 0.14722877 ₄.9680

Gender 1 1 0.00₄93087 0.166₄

Response Log IL-12p40 Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.3152₄616 7.9071 0.005₄ Age 1 1 0.0298₄605 0.7₄86 0.3879 Gender 1 1 0.00668055 0.1676 0.6827

Response Log IL-13 Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.17723383 ₄.07₄2 0.0₄₄8 Age 1 1 0.08598180 1.9765 0.1612 Gender 1 1 0.0065601₄ 0.1508 0.6982

Response Log IL-15 Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.12650₄20 2.6982 0.1019 Age 1 1 0.00₄5₄178 0.0969 0.7559 Gender 1 1 0.0017₄762 0.0373 0.8₄71

Response Log IL-16 Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.02167₄1₄ 1.1274

Age 1 1 0.14040023 7.3031

Gender 1 1 0.01948536 1.0136

Response Log IL-18 Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.00000090 0.0000

Age 1 1 0.02999650 0.9353

Gender 1 1 0.18201466 5.6754

Response Log IL-1 ra Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.00459148 0.0856

Age 1 1 0.00000672 0.0001

Gender 1 1 0.25180684 4.6921

Response Log IL-3 Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.30101880 4.2495

Age 1 1 0.11602487 1.6379

Gender 1 1 0.01103847 0.1558

Response Log IL-4 Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.11663309 2.9494

Age 1 1 0.00768277 0.1943

Gender 1 1 0.00092567 0.0234

Response Log IL-5 Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.11630491 2.8059 0.0954

Age 1 1 0.00001771 0.0004 0.9835

Gender 1 1 0.08699301 2.0987 0.1489 Response Log IL-6 Receptor Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.00000365 0.0002

Age 1 1 0.007₄5633 04033

Gender 1 1 0.000₄7136 0.0255

Response Log IL-7 Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.15411911 1.8845 0.1713 Age 1 1 0.22255314 2.7213 0.1005 Gender 1 1 0.00744424 0.0910 0.7632

Response Log IL-8 Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.04554019 1.9240 0.1669 Age 1 1 0.64728321 27.3474 <.0001 Gender 1 1 0.00614616 0.2597 0.6109

Response Log Insulin Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.03248195 0.2073 0.6493 Age 1 1 0.33195516 2.1190 0.1469 Gender 1 1 0.29256644 1.8676 0.1732

Response Log IP-10 Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.0032570 0.1000 0.7521 Age 1 1 1.2522396 38.4663 <.0001 Gender 1 1 0.1343761 4.1278 0.0434

Response Log Kidney Injury Molecule Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.02039153 0.2132 0.6447

Age 1 1 0.05990714 0.6263 0.4296

Gender 1 1 0.01620316 0.1694 0.6810

Response Log Leptin Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.1314386 0.8608 0.3546 Age 1 1 0.0101785 0.0667 0.7965 Gender 1 1 9.0073937 58.9869 <.0001

Response Log LH Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.02833032 0.5889 0.4437 Age 1 1 0.01620949 0.3369 0.5622 Gender 1 1 0.75698779 15.7349 <.0001

Response Log Lipoprotein (a) Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.1810974 0.5969

Age 1 1 1.1556530 3.8091

Gender 1 1 0.8659218 2.8541

Response Log MCP-1 Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.00477385 0.1678 0.6825

Age 1 1 0.00796370 0.2800 0.5973

Gender 1 1 0.12038512 4.2324 0.0409

Response Log MCP-2 Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.06396664 1.9093 0.1685

Age 1 1 0.13004599 3.8816 0.0501

Gender 1 1 0.00159426 0.0476 0.8275

Response Log MCP-3 Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.08305116 0.4915

Age 1 1 0.14898399 0.8817

Gender 1 1 0.01579239 0.0935 Response Log MCP-4 Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.35626731 2.8787 0.0912

Age 1 1 0.222₄025₄ 1.7971 0.1815

Gender 1 1 0.0750₄₄85 0.606₄ 0.₄370

Response Log M-CSF Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.003₄3069 0.5612 0.₄5₄6

Age 1 1 0.01211031 1.9812 0.1607

Gender 1 1 0.00₄03273 0.6597 0.₄176

Response Log MDC Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.11₄₄122_{4 4}.3831 0.0375

Age 1 1 0.02₄55237 0.9₄06 0.3332

Gender 1 1 0.1₄380126 5.5089 0.0198

Response Log MIF Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.002₄5775 0.035₄ 0.8509 Age 1 1 0.0001538₄ 0.0022 0.9625 Gender 1 1 0.09223797 1.3292 0.2502

Response Log Gamma Interferon Induced Monokine Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.1635090 2.26₄9 0.1338 Age 1 1 6.21₄1380 86.075₄ <.0001 Gender 1 1 0.0233275 0.3231 0.5703

Response Log MIP-1 alpha Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.01₄90893 0.7751 0.3796 Age 1 1 0.3561366₄ 18.5141 <.0001 Gender 1 1 0.00430177 0.2236 0.6368

Response Log MIP-1 beta Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.16751430 5.6689 0.0181 Age 1 1 0.28575218 9.6703 0.0021 Gender 1 1 0.00428598 0.1450 0.7037

Response Log MIP-3 alpha Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.01761420 0.3777 0.5395 Age 1 1 0.08578056 1.8392 0.1765 Gender 1 1 0.00882108 0.1891 0.6641

Response Log MMP-1 Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.09167418 1.8845 0.1713 Age 1 1 0.12919087 2.6558 0.1046 Gender 1 1 0.00451312 0.0928 0.7610

Response Log MMP-10 Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.04351991 2.2109 0.1385

Age 1 1 0.00903112 0.4588 0.4989

Gender 1 1 0.07561340 3.8413 0.0513

Response Log MMP-2 Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.01970485 1.7855 0.1829 Age 1 1 0.52235119 47.3301 <.0001 Gender 1 1 0.00027807 0.0252 0.8740

Response Log MMP-7 Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.49766396 8.2502 0.0045 Age 1 1 0.45035758 7.4660 0.0068 Gender 1 1 0.05850088 0.9698 0.3258 Response Log MMP9 (total) Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.01₄92707 0.3738 0.5₄16

Age 1 1 0.03₄1₄682 0.8550 0.3562

Gender 1 1 0.0012₄679 0.0312 0.8599

Response Log MMP-9 Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.00556767 0.106₄ 0.7₄₄6

Age 1 1 0.02₄6₄506 04710 0.₄933

Gender 1 1 0.00672152 0.1285 0.720₄

Response Log Myeloid Progenitor Inhibitory Factor 1 Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.00169608 0.0₄57 0.8310

Age 1 1 0.02595₄98 0.6991 0.₄0₄0

Gender 1 1 0.03353098 0.9032 0.3₄30

Response Log Myeloperoxidase Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.10312558 2.7195 0.1006

Age 1 1 0.03081056 0.8125 0.368₄

Gender 1 1 0.14757376 3.8917 0.0₄98

Response Log Myoglobin Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.0128826 0.2308 0.631₄

Age 1 1 0.9620369 17.2382 <.0001

Gender 1 1 1.5553831 27.8700 <.0001

Response Log NGAL Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.10265722 ₄.9015 0.0279

Age 1 1 0.12790852 6.1071 0.01₄2

Gender 1 1 0.0130₄5₄7 0.6229 0.₄309

Response Log NrCAM Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.02501126 0.8593 0.3550

Age 1 1 0.07681880 2.6393 0.1057

Gender 1 1 0.01927987 0.662₄ 0.₄166

Response Log NT-proBNP Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.329₄675 2.781₄ 0.0968

Age 1 1 7.59299₄0 64.1001 <.0001

Gender 1 1 0.4931799 4.1634 0.0425

Response Log Osteopontin Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.69657598 11.3035 0.0009

Age 1 1 0.41519074 6.7374 0.0101

Gender 1 1 0.00798441 0.1296 0.7192

Response Log PAI-1 Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.84859772 7.8749 0.0055

Age 1 1 0.44954157 4.1717 0.0423

Gender 1 1 0.08725773 0.8097 0.3692

Response Log Pancreatic Polypeptide Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.5332216 4.9389 0.0273

Age 1 1 1.1143541 10.3216 0.0015

Gender 1 1 0.4182280 3.8738 0.0503

Response Log Prostatic Acid Phosphatase Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.0070432 0.1998 0.6554

Age 1 1 0.0239748 0.6800 0.4105

Gender 1 1 1.7600504 49.9197 <.0001 Response Log PAPP-A Effect Tests

Source N arm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.029₄1113 0.9₄95 0.3309 Age 1 1 0.15681323 5.0628 0.0255 Gender 1 1 0.3976₄860 12.8382 0.000₄

Response Log PARC Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.00218937 0.200₄ 0.6548 Age 1 1 0.00189119 0.1731 0.6778 Gender 1 1 0.001238₄6 0.1134 0.7367

Response Log PDGF-BB Effect Tests

Source N arm DF Sum of Squares F Ratio

DIAG 1 1 2.1888090 12.0703

Age 1 1 0.6790357 3.7446

Gender 1 1 0.1269765 0.7002

Response Log PLGF Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.02038334 0.2180 0.6410 Age 1 1 0.05618906 0.6010 0.4390 Gender 1 1 0.42403864 4.5359 0.0343

Response Log Progesterone Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.1339600 1.4964 0.2226 Age 1 1 1.0648380 11.8946 0.0007 Gender 1 1 0.4020581 4.4911 0.0352

Response Log Proinsulin, Intact Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.0496975 0.4180 0.5186

Age 1 1 1.3717368 11.5379 0.0008

Gender 1 1 0.5645608 4.7486 0.0304

Response Log Proinsulin, Total Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.0041617 0.0368

Age 1 1 1.5782284 13.9460

Gender 1 1 0.4999796 4.4181

Response Log Prolactin Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.06380200 2.8963

Age 1 1 0.00159471 0.0724

Gender 1 1 0.07836238 3.5573

Response Clean Protein S Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 43.468069 3.6411

Age 1 1 25.415184 2.1289

Gender 1 1 5.438644 0.4556

Response Log PYY Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.17971507 2.0659 0.1521 Age 1 1 0.61364942 7.0541 0.0085 Gender 1 1 0.00082324 0.0095 0.9226

Response Log RANTES Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 1.0496024 5.3433 0.0218

Age 1 1 1.3382554 6.8128 0.0097

Gender 1 1 0.1576702 0.8027 0.3713

Response Log Resistin Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.27339616 10.7565 0.0012

Age 1 1 0.19935100 7.8432 0.0056

Gender 1 1 0.00751942 0.2958 0.5871 Response Clean Serum Amyloid P Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 106.0₄379 5.₄76₄ 0.0202

Age 1 1 1.80531 0.0932 0.760₄

Gender 1 1 87.55513 ₄.5216 0.03₄6

Response Log Stem Cell Factor Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.01852026 1.1148

Age 1 1 0.32018466 19.2730

Gender 1 1 0.01426403 0.8586

Response Log SGOT Effect Tests

Source N arm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.04847391 4.0115 0.0465 Age 1 1 0.02163685 1.7906 0.1823 Gender 1 1 0.00009816 0.0081 0.9283

Response Log SHBG Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.4339200 10.6601 0.0013 Age 1 1 0.5003729 12.2926 0.0006 Gender 1 1 1.2732352 31.2795 <.0001

Response Log SOD Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.00132145 0.0321 0.8579 Age 1 1 0.06223609 1.5132 0.2200 Gender 1 1 0.01496311 0.3638 0.5470

Response Log Sortilin

Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.04247027 4.6459 0.0322 Age 1 1 0.17688474 19.3498 <.0001 Gender 1 1 0.05852244 6.4019 0.0121

Response Log sRAGE Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.22861903 3.3367 0.0691 Age 1 1 0.00000033 0.0000 0.9982 Gender 1 1 0.07176324 1.0474 0.3073

Response Log Tamm-Horsfall Protein Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.00539457 0.1134 0.7367

Age 1 1 0.44474727 9.3467 0.0025

Gender 1 1 0.53750842 11.2962 0.0009

Response Log Thyroxine Binding Globulin Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.01656804 2.4247 0.1209 Age 1 1 0.00211486 0.3095 0.5786 Gender 1 1 0.09915053 14.5107 0.0002 Response Log TECK

Source Nparm Sum of Squares F Ratio Prob > F DIAG 1 0.00203065 0.0743 0.7854 Age 1 0.04045441 1.4807 0.2250 Gender 1 0.00243171 0.0890 0.7657

Response Log Tenasci Effect Tests

Source Nparm Sum of Squares F Ratio

DIAG 1 0.54092106 8.3190

Age 1 0.00836436 0.1286

Gender 1 0.06827918 1.0501

Response Log Testosterone Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.002196 0.0688 0.7933

Age 1 1 0.688003 21.5645 <.0001

Gender 1 1 12.640627 396.2021 <.0001 Response Log TGF-alpha Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.12985650 1.1246

Age 1 1 0.00027109 0.0023

Gender 1 1 0.57510231 4.9805

Response Log Thrombomodulin Effect Tests

Source N arm DF Sum of Squares F Ratio

DIAG 1 1 0.11910394 7.4531

Age 1 1 0.22303312 13.9566

Gender 1 1 0.08343555 5.2211

Response Log TIMP-1 Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.00448977 0.3429 0.5588 Age 1 1 0.31317984 23.9173 <.0001 Gender 1 1 0.05852278 4.4693 0.0357

Response Log TNFRII Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.05342657 2.3887 0.1237 Age 1 1 0.86126313 38.5072 <.0001 Gender 1 1 0.05332364 2.3841 0.1241

Response Log Thrombopoietin Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.01263842 0.5889

Age 1 1 0.06204650 2.8909

Gender 1 1 0.01252604 0.5836

Response Log TRAIL R3 Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.01891152 0.5502

Age 1 1 0.03540012 1.0299

Gender 1 1 0.01459360 0.4246

Response Clean Transferrin Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 36596.45 0.3276 0.5677 Age 1 1 651439.43 5.8307 0.0166 Gender 1 1 618910.26 5.5395 0.0195

Response Log Trefoil Factor 3 Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.0515857 1.5963 0.2078 Age 1 1 1.1693789 36.1866 <.0001 Gender 1 1 0.1108203 3.4293 0.0654

Response Log TSH Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.03925779 0.6584 0.4180 Age 1 1 0.00408833 0.0686 0.7937 Gender 1 1 0.01214189 0.2036 0.6523

Response Log Thrombospondin-1 Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 1.0240591 3.9154 0.0491 Age 1 1 2.0486098 7.8327 0.0056 Gender 1 1 0.0075214 0.0288 0.8655

Response Clean TTR Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 1575.074 0.5561 0.4567 Age 1 1 10815.801 3.8187 0.0520 Gender 1 1 42562.481 15.0274 0.0001

Response Log VCAM Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.03325608 2.5553 0.1114 Age 1 1 0.23542032 18.0892 <.0001 Gender 1 1 0.00586458 0.4506 0.5028 Response Log VEGF Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.072₄89₄2 2.8019 0.0956

Age 1 1 0.1793₄262 6.9321 0.0091

Gender 1 1 0.0₄0501₄3 1.5655 0.2122

Response Clean Vitronectin Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 99.155 0.0093 0.9232

Age 1 1 16652.525 1.5630 0.2126

Gender 1 1 6980.3₄7 0.6552 0.₄192

Response Log von Willebrand Factor Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.05169529 0.8901 0.3₄65

Age 1 1 0.556395₄6 9.5799 0.0022

Gender 1 1 0.0009₄01₄ 0.0162 0.8989

Table 9: UPenn Plasma Alzheimer's disease vs Other Univariate. Age and Gender

Response Log Alpha 1 -Antichymotrypsin Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.00309601 0.1352 0.7137

Age 1 1 0.01101623 04810 0.₄892

Gender 1 1 0.02913031 1.2719 0.261₄

Response Log Alpha-1 Antitrypsin Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.00003₄68 0.00₄9 0.9₄₄1

Age 1 1 0.00₄26070 0.6066 0.₄37₄

Gender 1 1 0.02097812 2.9866 0.0862

Response Log ACE Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.12788902 ₄.360₄ 0.0386

Age 1 1 0.01613932 0.5503 0.₄595

Gender 1 1 0.00000121 0.0000 0.99₄9

Response Log Adiponectin Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.103₄1838 1.6151 0.2059

Age 1 1 0.1₄212153 2.2196 0.1386

Gender 1 1 045887861 7.1666 0.0083

Response Log Agouti Related Protein Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.0330₄685 04656 04962

Age 1 1 0.28697202 ₄.0₄28 0.0₄63

Gender 1 1 0.6636₄60₄ 9.3₄92 0.0027

Response Log Alpha-2 Macroglobulin Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.0₄₄19₄61 9.2₄95 0.0028

Age 1 1 0.00086357 0.1807 0.671₄

Gender 1 1 0.02327618 ₄.8715 0.0290

Response Log Alpha-Fetoprotein Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.0089983 0.1033 0.7₄8₄

Age 1 1 1.5809929 18.1484 <.0001

Gender 1 1 1.0699383 12.2820 0.0006

Response Log Alphal Microglobulin Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F Source N arm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.00008132 0.00₄2 0.9₄8₄ Age 1 1 0.06719291 34673 0.06₄8 Gender 1 1 0.003550₄8 0.1832 0.6693

Response Log ANG-2 Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.062₄7723 2.52₄6

Age 1 1 0.20147025 8.1412

Gender 1 1 0.05364905 2.1679

Response Log Angiotensinogen Effect Tests

Source N arm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.00390611 0.0053 0.9419 Age 1 1 0.17683485 0.2416 0.6239 Gender 1 1 0.02410816 0.0329 0.8563

Response Log Apo A1 Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.35398992 14.3085

Age 1 1 0.00227715 0.0920

Gender 1 1 0.23558010 9.5223

Response Log Apo A2 Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.00054261 0.0457

Age 1 1 0.01237080 1.0419

Gender 1 1 0.09003143 7.5823

Response Log ApoCI Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.00457215 0.3685

Age 1 1 0.00234256 0.1888

Gender 1 1 0.14493017 11.6798

Response Log Apo Clll Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.03182938 1.4841

Age 1 1 0.05736538 2.6748

Gender 1 1 0.52922429 24.6765

Response Log ApoH Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.00000022 0.0000 0.9959 Age 1 1 0.00163374 0.1942 0.6601 Gender 1 1 0.00151648 0.1803 0.6718

Response Log Apo A-IV Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.01198971 0.2649 0.6076 Age 1 1 0.01298367 0.2869 0.5931 Gender 1 1 0.00250276 0.0553 0.8144

Response Log ApoB Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.00747490 0.5144

Age 1 1 0.01068622 0.7353

Gender 1 1 0.03065425 2.1094

Response Log ApoD Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.27842576 15.3916 0.0001

Age 1 1 0.00140484 0.0777 0.7809

Gender 1 1 0.29054201 16.0614 0.0001

Response Log ApoE Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.15412082 4.3330 0.0393

Age 1 1 0.01895254 0.5328 0.4667

Gender 1 1 0.32049486 9.0106 0.0032 Response Log ApoJ Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.00076333 0.1500

Age 1 1 0.00105723 0.2077

Gender 1 1 0.072₄3679 1₄.2298

Response Log AXL Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.06₄08687 3.1370

Age 1 1 0.000₄2900 0.0210

Gender 1 1 0.00318₄7₄ 0.1559

Response Log Beta2 Microglobulin Effect Tests

Source N arm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.00018₄30 0.0111 0.916₄ Age 1 1 0.21665283 12.9922 0.000₄ Gender 1 1 0.00228699 0.1371 0.7117

Response Log BLC Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.1328₄888 2.03₄9

Age 1 1 0.19277850 2.9528

Gender 1 1 0.00007850 0.0012

Response Log BMP -6 Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.01137700 0.1279 0.7212

Age 1 1 0.₄0209₄9_{4 4}.519₄ 0.0353

Gender 1 1 0.05096163 0.5728 0.₄505

Response Log BDNF Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 5.8863603 26.97₄5 <.0001

Age 1 1 1.2370992 5.6691 0.0187

Gender 1 1 0.12853₄1 0.5890 0.₄₄₄1

Response Log Betacellulin Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.₄₄6₄₄167 54089

Age 1 1 0.00₄₄6691 0.05₄1

Gender 1 1 0.2₄531₄32 2.9721

Response Clean C3 Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.07107676 3.1372

Age 1 1 0.00130196 0.0575

Gender 1 1 0.19615605 8.6579

Response Log Cancer Antigen 19-9 Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.00010009 0.0002 0.9883 Age 1 1 0.1₄507591 0.3135 0.5765 Gender 1 1 0.316733₄₄ 0.68₄₄ 0.₄095

Response Log CD40 Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.00067₄06 0.0311 0.8602

Age 1 1 0.07099809 3.2786 0.0724

Gender 1 1 0.0269₄8₄₄ 1.2444 0.2666

Response Log CD40 Ligand Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 5.3333244 45.8487 <.0001 Age 1 1 0.8491355 7.2997 0.0078 Gender 1 1 0.0041466 0.0356 0.8505

Response Log CDL5 Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.07212531 2.0438 0.1551

Age 1 1 0.18976651 5.3773 0.0219

Gender 1 1 0.02635455 0.7468 0.3890 Response Log Carcinoembryonic Antigen Effect Tests

Source N arm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.255222₄ 3.823₄ 0.0526 Age 1 1 1.930₄8₄8 28.9197 <.0001 Gender 1 1 0.0010399 0.0156 0.9009

Response Log CgA Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.03239612 0.5005

Age 1 1 0.99289559 15.3401

Gender 1 1 0.01869886 0.2889

Response Log CKMB Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.2927776 4.9816

Age 1 1 0.3491461 5.9407

Gender 1 1 1.0315442 17.5517

Response Log CNTF Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.08330005 1.1829

Age 1 1 0.00113542 0.0161

Gender 1 1 0.00605224 0.0859

Response Log Complement Factor H Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.00430038 0.1146 0.7355 Age 1 1 0.00018750 0.0050 0.9437 Gender 1 1 0.02437967 0.6498 0.4216

Response Log CTGF Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.00043029 0.0083 0.9274 Age 1 1 0.04533892 0.8770 0.3507 Gender 1 1 0.00233893 0.0452 0.8319

Response Log Cortisol Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.03407374 2.3088

Age 1 1 0.00588324 0.3986

Gender 1 1 0.06453539 4.3729

Response Log C-Peptide Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.40096318 5.2768

Age 1 1 0.47157244 6.2061

Gender 1 1 0.52455378 6.9033

Response Log CRP Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.69684447 2.1059

Age 1 1 0.17892658 0.5407

Gender 1 1 0.95170846 2.8761

Response Log Cystatin Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.00149139 0.1000 0.7524

Age 1 1 0.08452885 5.6661 0.0187

Gender 1 1 0.00387165 0.2595 0.6113

Response Log EGF Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 5.7831540 43.1596 <.0001

Age 1 1 1.2021661 8.9717 0.0033

Gender 1 1 0.0127866 0.0954 0.7579

Response Log EGF-R Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.07862613 4.6074

Age 1 1 0.01194499 0.7000

Gender 1 1 0.00419388 0.2458 Response Log ENA-78 Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 3.5₄73235 27.0031

Age 1 1 0.8711928 6.6317

Gender 1 1 0.0₄26209 0.32₄₄

Response Log Endothelin-1 Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.02088391 04172

Age 1 1 0.00802₄19 0.1603

Gender 1 1 0.613375₄0 12.2523

Response Log EN-RAGE Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.00280₄38 0.0287

Age 1 1 0.00853931 0.0873

Gender 1 1 0.02036726 0.2082

Response Log Eotaxin Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.38195213 5.33₄7

Age 1 1 0.13505149 1.8863

Gender 1 1 0.00712569 0.0995

Response Log E-Selectin Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.01933584 0.4714 0.4935 Age 1 1 0.00225549 0.0550 0.8150 Gender 1 1 0.01286026 0.3135 0.5764

Response Log FABP Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.0155896 0.0720 0.7889 Age 1 1 1.0997873 5.0796 0.0258 Gender 1 1 1.1925547 5.5080 0.0204

Response Clean Factor VII Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 29923.76 0.7733 0.3808

Age 1 1 2673.44 0.0691 0.7931

Gender 1 1 640558.83 16.5528 <.0001 Response Log FAS Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.02418242 1.5772

Age 1 1 0.12992281 8.4735

Gender 1 1 0.09010678 5.8767

Response Log Fas-Ligand Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.04207548 1.0085

Age 1 1 0.13578602 3.2545

Gender 1 1 0.00454046 0.1088

Response Log Ferritin Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.05577516 0.3341 0.5642

Age 1 1 0.28652951 1.7164 0.1924

Gender 1 1 0.94816993 5.6798 0.0185

Response Log Fetuin A Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.00004587 0.0062 0.9373

Age 1 1 0.00173568 0.2353 0.6284

Gender 1 1 0.04963110 6.7278 0.0105

Response Log Fibrinogen Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.00096977 0.0680

Age 1 1 0.00082767 0.0581

Gender 1 1 0.04520213 3.1707 Response Log FSH Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.0180₄88 04863 0.₄868

Age 1 1 0.2073365 5.5865 0.0195

Gender 1 1 ₄.7₄20₄35 127.7710 <.0001

Response Log G-CSF Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.00196731 0.0₄59

Age 1 1 0.00062115 0.01₄5

Gender 1 1 0.30788138 7.1823

Response Log Growth Hormone Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.02155816 0.0₄57

Age 1 1 0.51065071 1.0830

Gender 1 1 0.6002₄839 1.2730

Response Log GLP-1 Total Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.02058₄₄6 0.8561

Age 1 1 0.006738₄0 0.2803

Gender 1 1 0.15567323 6.₄7₄₄

Response Log Glucagon Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.0₄7689₄5 1.2788 0.2601

Age 1 1 0.11411257 3.0599 0.0825

Gender 1 1 0.126₄690₄ 3.3912 0.0677

Response Log GST alpha Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.27099280 1.7714 0.1854

Age 1 1 0.02735270 0.1788 0.6731

Gender 1 1 0.22099099 14446 0.2315

Response Log GRO-alpha Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 2.7151783 30.4431

Age 1 1 0.2560178 2.8705

Gender 1 1 0.0549920 0.6166

Response Log Haptoglobul Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.98291792 3.1142

Age 1 1 0.04589758 0.1454

Gender 1 1 0.28813856 0.9129

Response Log HB-EGF Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 3.6513564 17.7594 <.0001 Age 1 1 0.1112133 0.5409 0.4633 Gender 1 1 0.0633231 0.3080 0.5798

Response Log HCC-4 Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.00026551 0.0049 0.9443

Age 1 1 0.08853004 1.6364 0.2030

Gender 1 1 0.03163332 0.5847 0.4458

Response Log HGF Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.02700502 2.1107 0.1486

Age 1 1 0.00008849 0.0069 0.9338

Gender 1 1 0.03155367 2.4663 0.1186

Response Log HSP60 Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.19081158 2.6162 0.1081

Age 1 1 0.00130405 0.0179 0.8938

Gender 1 1 0.27482413 3.7680 0.0543 Response Log 1-309 Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.0032₄637 0.0121

Age 1 1 0.61₄36763 2.2903

Gender 1 1 0.27₄17915 1.0221

Response Log ICAM-1 Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.06₄60391 3.3366

Age 1 1 0.06765591 3.₄9₄3

Gender 1 1 0.09₄709₄6 ₄.8915

Response Log IgA Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.00560051 0.1144

Age 1 1 0.13078951 2.6709

Gender 1 1 0.00389298 0.0795

Response Log IgE Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.2880261 0.7577

Age 1 1 0.5265826 1.3852

Gender 1 1 3.7779222 9.9382

Response Log IGFBP-2 Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.02609120 0.4563

Age 1 1 0.45355998 7.9321

Gender 1 1 0.00496559 0.0868

Response Log IgM Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.04135207 0.7445

Age 1 1 0.02000818 0.3602

Gender 1 1 0.02618176 0.4714

Response Log IL-10 Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.03805695 1.2132

Age 1 1 0.06472602 2.0633

Gender 1 1 0.00351746 0.1121

Response Log IL-12p40 Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.10374307 2.4214 0.1220 Age 1 1 0.00768044 0.1793 0.6727 Gender 1 1 0.01550620 0.3619 0.5484

Response Log IL-13 Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.48911995 10.7788

Age 1 1 0.04455125 0.9818

Gender 1 1 0.08630200 1.9018

Response Log IL-15 Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.06828949 1.6463

Age 1 1 0.06170346 1.4876

Gender 1 1 0.00943827 0.2275

Response Log IL-16 Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.00020927 0.0094

Age 1 1 0.01403645 0.6285

Gender 1 1 0.01030902 0.4616

Response Log IL-18 Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.06469431 2.1718

Age 1 1 0.01594176 0.5352

Gender 1 1 0.03580745 1.2021 Response Log IL-1 ra Effect Tests

Source N arm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.6695330₄ 11.2399 0.0010 Age 1 1 0.001813₄₄ 0.030₄ 0.8617 Gender 1 1 0.05₄15893 0.9092 0.3₄20

Response Log IL-3 Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.12863518 1.7065 0.1936 Age 1 1 0.27183199 3.6063 0.0597 Gender 1 1 0.076₄₄101 1.0141 0.3157

Response Log IL-4 Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.08454037 2.5767 0.1108 Age 1 1 0.02059001 0.6276 0.4296 Gender 1 1 0.03871825 1.1801 0.2793

Response Log IL-5 Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.20577596 5.4678 0.0208 Age 1 1 0.01898051 0.5043 0.4788 Gender 1 1 0.05584317 1.4839 0.2253

Response Log IL-6 Receptor Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.00029649 0.0168

Age 1 1 0.00328296 0.1856

Gender 1 1 0.00095843 0.0542

Response Log IL-7 Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.80119996 8.9895 0.0032

Age 1 1 0.14740498 1.6539 0.2006

Gender 1 1 0.00790063 0.0886 0.7664

Response Log IL-8 Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.03424653 1.2286 0.2696 Age 1 1 0.16540503 5.9341 0.0161 Gender 1 1 0.08756335 3.1415 0.0786

Response Log Insulin Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.77207491 5.2419 0.0236 Age 1 1 0.32339507 2.1956 0.1407 Gender 1 1 0.76597402 5.2004 0.0241

Response Log IP-10 Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.47718825 11.2782 0.0010 Age 1 1 0.99979975 23.6300 <.0001 Gender 1 1 0.20338323 4.8069 0.0300

Response Log Kidney Injury Molecule Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.24540713 2.0287 0.1566 Age 1 1 0.38962471 3.2209 0.0749 Gender 1 1 0.21381579 1.7676 0.1859

Response Log Leptin Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.2102846 1.1542

Age 1 1 0.0559775 0.3072

Gender 1 1 5.6205406 30.8490

Response Log LH Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.03608443 0.9000 0.3445 Age 1 1 0.02487082 0.6203 0.4323 Gender 1 1 0.70077750 17.4793 <.0001 Response Log Lipoprotein (a) Effect Tests

Source N arm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.6₄0106₄8 1.8548 0.1755 Age 1 1 0.01158812 0.0336 0.8549 Gender 1 1 0.33952815 0.9838 0.3230

Response Log MCP-1 Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.13273292 4.3513 0.0388

Age 1 1 0.00248396 0.0814 0.7758

Gender 1 1 0.01589563 0.5211 0.4716

Response Log MCP-2 Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.56024410 14.1756 0.0002

Age 1 1 0.02406649 0.6089 0.4365

Gender 1 1 0.01171750 0.2965 0.5870

Response Log MCP-3 Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.22430481 1.6098 0.2067

Age 1 1 0.01148317 0.0824 0.7745

Gender 1 1 0.15744123 1.1299 0.2897

Response Log MCP-4 Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 1.3380887 13.1311 0.0004

Age 1 1 0.1033455 1.0142 0.3157

Gender 1 1 0.0069750 0.0684 0.7940

Response Log M-CSF Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.01223382 1.6048 0.2074

Age 1 1 0.00164693 0.2160 0.6428

Gender 1 1 0.00317576 0.4166 0.5197

Response Log MDC Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.00142175 0.0647 0.7996

Age 1 1 0.02180147 0.9917 0.3211

Gender 1 1 0.06813425 3.0994 0.0806 Response Log MIF Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.05392112 0.7212 0.3973

Age 1 1 0.20658690 2.7630 0.0988

Gender 1 1 0.00458773 0.0614 0.8047

Response Log Gamma Interferon Induced Monokine Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.1217418 1.6536 0.2007 Age 1 1 2.7416123 37.2395 <.0001 Gender 1 1 0.4695848 6.3784 0.0127

Response Log MIP-1 alpha Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.04013824 2.2937 0.1322 Age 1 1 0.10221209 5.8408 0.0170 Gender 1 1 0.04973944 2.8423 0.0941

Response Log MIP-1 beta Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.35621626 13.3533 0.0004 Age 1 1 0.22454016 8.4172 0.0043 Gender 1 1 0.00009980 0.0037 0.9513

Response Log MIP-3 alpha Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.07864725 1.5497 0.2153 Age 1 1 0.07718807 1.5210 0.2196 Gender 1 1 0.00208289 0.0410 0.8398 Response Log MMP-1 Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.37703553 10.1300 0.0018

Age 1 1 0.0011₄50₄ 0.0308 0.8610

Gender 1 1 0.032850₄1 0.8826 0.3₄92

Response Log MMP-10 Effect Tests

Source N arm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.0₄0875₄₄ 2.1925 0.1410 Age 1 1 0.0720₄161 3.8642 0.0514 Gender 1 1 0.03161756 1.6959 0.1950

Response Log MMP-2 Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.00549654 0.4375 0.5095

Age 1 1 0.46914122 37.3392 <.0001

Gender 1 1 0.01350779 1.0751 0.3016

Response Log MMP-7 Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.00213612 0.0363 0.8492

Age 1 1 0.25108745 4.2659 0.0408

Gender 1 1 0.06678099 1.1346 0.2887

Response Log MMP9 (total) Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.02118275 0.4291 0.5135 Age 1 1 0.02350467 0.4762 0.4913 Gender 1 1 0.00014502 0.0029 0.9569

Response Log MMP-9 Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.10829191 1.8615 0.1747 Age 1 1 0.00677530 0.1165 0.7334 Gender 1 1 0.00643394 0.1106 0.7400

Response Log Myeloid Progenitor Inhibitory Factor 1 Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.10586132 2.3696 0.1260 Age 1 1 0.04753259 1.0640 0.3041 Gender 1 1 0.08751542 1.9590 0.1639 Response Log Myeloperoxidase Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.32541400 6.2139 0.0139 Age 1 1 0.00573308 0.1095 0.7413 Gender 1 1 0.04153126 0.7931 0.3747

Response Log Myoglobin Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.0572154 1.0335 0.3112 Age 1 1 0.0748181 1.3514 0.2471 Gender 1 1 1.1977325 21.6340 <.0001

Response Log NGAL Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.00005946 0.0026 0.9591 Age 1 1 0.00040492 0.0180 0.8935 Gender 1 1 0.04356045 1.9359 0.1664

Response Log NrCAM Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.05176007 1.7843 0.1839 Age 1 1 0.00168820 0.0582 0.8097 Gender 1 1 0.02007147 0.6919 0.4070

Response Log NT-proBNP Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.0713455 0.5140 0.4746

Age 1 1 3.4520659 24.8705 <.0001

Gender 1 1 0.5734034 4.1311 0.0440 Response Log Osteopontin Effect Tests

Source N arm DF Sum of Squares F Ratio

DIAG 1 1 0.0₄777₄78 0.7799

Age 1 1 0.15210632 24829

Gender 1 1 0.00637867 0.1041

Response Log PAI-1 Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 3.1379704 28.8223

Age 1 1 0.0378878 0.3480

Gender 1 1 0.0344729 0.3166

Response Log Pancreatic Polypeptide Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.16879415 1.3107 0.2543

Age 1 1 0.11647147 0.9044 0.3433

Gender 1 1 0.34910642 2.7107 0.1020

Response Log Prostatic Acid Phosphatase Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.0503131 1.3709 0.2437

Age 1 1 0.1821020 4.9617 0.0276

Gender 1 1 1.7033483 46.4106 <.0001

Response Log PAPP-A Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.03550076 1.1853 0.2782 Age 1 1 0.03452393 1.1527 0.2849 Gender 1 1 0.41685441 13.9179 0.0003

Response Log PARC Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.00110612 0.0950 0.7584 Age 1 1 0.01737447 1.4921 0.2240 Gender 1 1 0.00000167 0.0001 0.9904

Response Log PDGF-BB Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 5.1181302 26.8029

Age 1 1 0.6215571 3.2550

Gender 1 1 0.0351862 0.1843

Response Log PLGF Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 1.5104813 18.1755 <.0001 Age 1 1 0.0047919 0.0577 0.8106 Gender 1 1 0.0152165 0.1831 0.6694

Response Log Progesterone Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.63537691 6.8566

Age 1 1 0.70500886 7.6080

Gender 1 1 0.79590559 8.5889

Response Log Proinsulin, Intact Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.30662306 3.2351

Age 1 1 0.63212507 6.6695

Gender 1 1 0.32759868 3.4564

Response Log Proinsulin, Total Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.35606379 3.4115

Age 1 1 0.52889525 5.0674

Gender 1 1 0.40346671 3.8656

Response Log Prolactin Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.02522257 0.9493 0.3316

Age 1 1 0.00001936 0.0007 0.9785

Gender 1 1 0.16903823 6.3623 0.0128 Response Clean Protein S Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.037866 0.0029

Age 1 1 0.₄8000₄ 0.0363

Gender 1 1 22.59265₄ 1.7082

Response Log PYY Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.20813797 3.2375 0.07₄2 Age 1 1 0.15276993 2.3763 0.1255 Gender 1 1 0.0360₄158 0.5606 0.₄553

Response Log RANTES Effect Tests

Source N arm DF Sum of Squares F Ratio Prob > F DIAG 1 1 5.162₄809 26.9910 <.0001 Age 1 1 1.0057119 5.2582 0.023₄ Gender 1 1 0.0699256 0.3656 0.5₄6₄

Response Log Resistin Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.03381765 1.2303

Age 1 1 0.0₄698379 1.7093

Gender 1 1 0.0₄391738 1.5977

Response Clean Serum Amyloid P Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 11.017618 0.₄9₄8 0.4830 Age 1 1 9.088078 04082 0.5240 Gender 1 1 41.314326 1.8555 0.1754

Response Log Stem Cell Factor Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.20542916 9.4332 0.0026

Age 1 1 0.00215016 0.0987 0.7538

Gender 1 1 0.00232134 0.1066 0.7446

Response Log SGOT Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.00248217 0.2476 0.6196 Age 1 1 0.00000158 0.0002 0.9900 Gender 1 1 0.00524082 0.5228 0.4709

Response Log SHBG Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.12091206 2.8407 0.0942 Age 1 1 0.05331419 1.2526 0.2650 Gender 1 1 0.67269330 15.8043 0.0001

Response Log SOD Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.03006726 0.6490

Age 1 1 0.09787141 2.1126

Gender 1 1 0.00016725 0.0036

Response Log Sortilin Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.09242051 9.9103 0.0020

Age 1 1 0.04792720 5.1393 0.0250

Gender 1 1 0.03575330 3.8339 0.0523

Response Log sRAGE Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.02382189 0.4164 0.5198

Age 1 1 0.00079873 0.0140 0.9061

Gender 1 1 0.14626676 2.5567 0.1121

Response Log Tamm-Horsfall Protein Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.11212570 2.1025 0.1494

Age 1 1 0.16144443 3.0274 0.0841

Gender 1 1 0.13721515 2.5730 0.1110 Response Log Thyroxine Binding Globulin Effect Tests

Source N arm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.0022₄600 0.3218 0.5715 Age 1 1 0.00085932 0.1231 0.7262 Gender 1 1 0.01889077 2.7066 0.1022

Response Log TECK Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.19896132 94176

Age 1 1 0.002₄78₄1 0.1173

Gender 1 1 0.00826751 0.3913

Response Log Tenasci Effect Tests

Source Nparm Sum of Squares F Ratio

DIAG 1 0.0517₄₄32 0.8593

Age 1 0.15720199 2.6106

Gender 1 0.001₄0192 0.0233

Response Log Testosterone Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.0186363 0.5271

Age 1 1 0.17625₄1 ₄.9855

Gender 1 1 7.09829₄9 200.7807

Response Log TGF-alpha Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 1.0100467 9.1677 0.0029

Age 1 1 0.0073251 0.0665 0.7969

Gender 1 1 0.0932355 0.8463 0.3592

Response Log Thrombomodulin Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F DIAG 1 1 0.01356193 0.9942 0.3205 Age 1 1 0.00477802 0.3503 0.5549 Gender 1 1 0.05486106 4.0219 0.0469

Response Log TIMP-1 Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.08152942 6.3472

Age 1 1 0.09602086 7.4753

Gender 1 1 0.00179554 0.1398

Response Log TNFRII Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.00335240 0.1355

Age 1 1 0.31227349 12.6252

Gender 1 1 0.00108792 0.0440

Response Log Thrombopoietin Effect Tests

Source Nparm DF Sum of Squares F Ratio

DIAG 1 1 0.07698395 2.9021

Age 1 1 0.17209602 6.4875

Gender 1 1 0.02097030 0.7905

Response Log TRAIL R3 Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.01820380 0.5790 0.4480

Age 1 1 0.00437378 0.1391 0.7097

Gender 1 1 0.08117603 2.5821 0.1104

Response Clean Transferrin Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 288181.84 2.4253 0.1217

Age 1 1 325852.38 2.7424 0.1000

Gender 1 1 365861.43 3.0791 0.0816

Response Log Trefoil Factor 3 Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.00023729 0.0073 0.9322

Age 1 1 0.32991606 10.1142 0.0018

Gender 1 1 0.10006665 3.0677 0.0821 Response Log TSH Effect Tests

Source N arm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.10070560 1.3326 0.2504

Age 1 1 0.0992789₄ 1.3138 0.2537

Gender 1 1 0.02936577 0.3886 0.5341

Response Log Thrombospondin-1 Effect Tests

Source N arm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 5.6₄03855 25.3429 <.0001

Age 1 1 1.4831393 6.6639 0.0109

Gender 1 1 0.1794649 0.8064 0.3708

Response Clean TTR Effect Tests

Source N arm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 179.5341 0.0675 0.7954

Age 1 1 5044.5965 1.8963 0.1708

Gender 1 1 6300.4605 2.3684 0.1261

Response Log VCAM Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.00441269 0.5385 0.4643

Age 1 1 0.14738921 17.9860 <.0001

Gender 1 1 0.00541127 0.6603 0.4179

Response Log VEGF Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.02429722 0.8876 0.3478

Age 1 1 0.04232906 1.5463 0.2158

Gender 1 1 0.00289003 0.1056 0.7457

Response Clean Vitronectin Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 7.550 0.0007 0.9782

Age 1 1 16059.943 1.5905 0.2094

Gender 1 1 792.248 0.0785 0.7798

Response Log von Willebrand Factor Effect Tests

Source Nparm DF Sum of Squares F Ratio Prob > F

DIAG 1 1 0.33853498 5.6042 0.0193

Age 1 1 0.38269567 6.3352 0.0130

Gender 1 1 0.04314735 0.7143 0.3995

For purposes of the univariate analysis, analytes that had more than 10% missing values were excluded from the analysis. Values that were reported as LOW (below the lowest assay limit which is defined by RBM as a value below the lowest calibrator for the individual assay) were imputed by taking the reported low detection limit for individual assays (LDD) and dividing by 2. Individual analyte distributions were tested for normalcy using Box-Cox analysis. If necessary, analytes were log transformed and checked again for normalcy. The majority of analytes exhibited non-normal distributions and were log transformed. For purposes of the univariate analysis, multidimensional scaling and Mahalanobis distances were used to detect outliers. Outliers were defined as 5 standard deviations beyond the mean and were replaced with the 5 STD value. All Univariate results corrected for age and gender. A separate analysis also corrected for age, gender and ApoE. Table 10 shows Univariate results from UPenn Plasma Dataset. The results demonstrate that levels of some biomarkers were increased compared to a reference level. Whereas levels of other biomarkers were decreased compared to a reference level.

Table 10:

Analytes Analyte Analyte

Alpha 1 Antichymotrypsin† Fibrinogen† Pancreatic Polypeptide † 1

Alpha 1- Antitrypsin† G-CSF † PAI-I i

Alpha 2 Macroglobulin † GST i PDGF-BB i

Apolipoprotein H† GRO-alpha j Protein S†

Apolipoprotein D† 1-309† RANTES i

Apolipoprotein E J, IGF-BP2† Resistin†

Betacelluli n J, IL12p40 i Serum Amyloid P J,

CD40† IL-8† SGOT i

CD40 Ligand j IL-10 † SHBG†

CRP i IL-13† Sortilin J,

CKMB i IL-3† Tenascin C†

Cortisol† IgA† Thrombomodulin †

Cystatin C † MDC †

EGF i MIP1 beta j

EN-RAGE † MMP-7†

E-Selectin J, NGAL †

FAS† Osteopontin†

Table 1 1 shows Univariate results from UPenn CSF Dataset. The results demonstrate that levels of some biomarkers were increased compared to a reference level. Whereas levels of other biomarkers were decreased compared to a reference level.

Table 1 1 :

Fibrinogen J,

IGF-BP2†

Table 12 summarizes a performance model of AD versus Control with respect to the 24 top plasma analytes. Table 12 shows that some markers are increased and other markers are decreased as wells as how each marker is altered between plasma and CSF samples. It was also observed that certain markers correlate with other markers to provide an indication of disease state.

Table 12:

The disclosures of each and every patent, patent application, and publication cited herein are hereby incorporated herein by reference in their entirety.

While the invention has been disclosed with reference to specific embodiments, it is apparent that other embodiments and variations of this invention may be devised by others skilled in the art without departing from the true spirit and scope of the invention. The appended claims are intended to be construed to include all such embodiments and equivalent variations.

Claims

CLAIMS What is claimed is:

1. A composition for detecting Alzheimer's Disease in an individual with at least 80% sensitivity, comprising reagents for detecting a plurality of biomarkers for

Alzheimer's Disease in a biological sample from the individual.

2. The composition of claim 1 capable of detecting biomarkers of Alzheimer's disease in a plasma sample.

3. The composition of claim 1, wherein said biomarkers for Alzheimer's disease comprise Cortisol, pancreatic polypeptide, osteopontin, IGF BP2, and resistin.

4. The composition of claim 1, wherein said biomarkers for Alzheimer's disease comprise resistin, pancreatic polypeptide, ApoD, G-CSF, MlPlbeta, ApoE,

NrCAM, MMP2, Cortisol, PDGF-BB, MMP1, and 1-309.

5. The composition of claim 1, wherein said biomarkers for Alzheimer's disease comprise Alpha- 1 Microglobulin, Angiopoietin-2, Apolipoprotein E, Beta-2 Microglobulin, Chemokine (C-X-C motif) ligand 13 (CXCL13) also known as B lymphocyte chemoattractant (BLC), Cortisol, E-Selectin, FAS, Fatty Acid Binding Protein, Hepatocyte Growth Factor, IGF BP-2, Interleukin 10, NT-Pro-BNP (Brain natriuretic peptide co-secreted along with a 76 amino acid N-terminal fragment), Osteopontin, Pancreatic Polypeptide, PAPP-A, Resistin, Stem Cell Factor, Tenascin C, Thrombomodulin, tissue inhibitor of metalloproteinases-1 (TIMP-1), vascular cell adhesion molecule- 1 (VCAM-1), Vascular endothelial growth factor (VEGF), and Von Willebrand Factor.

6. The composition of claim 1, wherein said biomarkers is one or more biomarkers selected from Table 2.

7. The composition of claim 1, wherein the reagent is an antibody.

8. The composition of claim 1 or claim 7, wherein the reagent is on a solid support.

9. A method of diagnosing Alzheimer's Disease in an individual, the method comprising analyzing a biological sample from an individual to determine the level(s) of one or more biomarkers for Alzheimer's Disease in the sample, wherein the one or more biomarkers are selected from Table 2; and comparing the level(s) of the one or more biomarkers in the sample to Alzheimer's Disease-positive and/or Alzheimer's Disease-negative reference levels of the one or more biomarkers in order to diagnose whether the individual has Alzheimer's Disease.

10. The method of claim 9, further comprising analyzing age, gender, and ApoE genotype (e2/e2, e2/e3, e2/e4, e3/e3, e3/e4 and e4/e4) of the individual in order to diagnose whether the individual has Alzheimer's Disease.

11. The method of claim 10, wherein the levels of the biomarkers, age, gender, and ApoE genotype are incorporated into an algorithm to diagnose whether the individual has Alzheimer's Disease.

12. The method of claim 1 1, wherein said algorithm is selected from the group consisting of Logistic Regression, Boosted Tree Models, Flexible Discriminant Analysis (FDA), K-Nearest Neighbors (KNN), Naive Bayes, Partial Least Squares (PLS), Randon Forests, Shrunken Centroids, Sparse Partial Least Squares, Support Vector Machines approaches, and any combination thereof.

13. The method of claim 9, wherein Alzheimer's Disease diagnosed is selected from the group consisting of late onset Alzheimer's disease, early onset

Alzheimer's disease, familial Alzheimer's disease, and sporadic Alzheimer's disease.

14. The method of claim 9, wherein said biological sample is selected from the group consisting of whole blood, a blood component, CSF, urine, and any combination thereof.

15. A method for determining whether dementia in an individual is associated with AD, the method comprising analyzing a biological sample from an individual to determine the level(s) of Cortisol, pancreatic polypeptide, osteopontin, IGF BP2, and resistin in the sample, and comparing the level(s) of Cortisol, pancreatic polypeptide, osteopontin, IGF BP-2, and resistin in a corresponding reference sample to determine whether dementia in an individual is associated with AD.

16. The method of claim 15, wherein the dementia is associated with AD.

17. The method of claim 15 or 16, further comprising analyzing age, gender, and ApoE genotype (e2/e2, e2/e3, e2/e4, e3/e3, e3/e4 and e4/e4) of the individual in order to determine the presence or status of dementia in the individual.

18. The method of claim 16, wherein the biological sample from the individual is analyzed to determine the level(s) of one or more biomarkers selected from Table 2; and comparing the level(s) of the one or more biomarkers in the sample to Alzheimer's Disease-positive and/or Alzheimer's Disease-negative reference levels of the one or more biomarkers in order to determine the presence or status of Alzheimer's Disease or other types of dementia in the individual.

19. The method of claim 18, wherein the levels of Cortisol, pancreatic polypeptide, osteopontin, IGF BP2, resistin biomarkers, age, gender, and ApoE genotype are incorporated into an algorithm in order to determine the presence or status of

Alzheimer's Disease or other types of dementia in the individual.

20. The method of claim 19, wherein said algorithm is selected from the group consisting of Logistic Regression, Boosted Tree Models, Flexible Discriminant Analysis (FDA), K-Nearest Neighbors (KNN), Naive Bayes, Partial Least Squares (PLS), Randon Forests, Shrunken Centroids, Sparse Partial Least Squares, Support Vector Machines approaches, and any combination thereof.

21. A method for diagnosing Alzheimer's disease in an individual, the method comprising: measuring the level of one or more biomarkers selected from Table 2 in a plasma sample from the individual, and comparing the level of the biomarkers in the plasma sample of the individual to positive control reference values and optionally to negative control reference values, wherein levels of the biomarkers in the plasma sample of the individual resemble the positive control reference values and do not resemble the negative control reference values are an indication that said individual has Alzheimer's disease.

22. The method of claim 21, further comprising analyzing age, gender, and ApoE genotype (e2/e2, e2/e3, e2/e4, e3/e3, e3/e4 and e4/e4) of the test individual in order to diagnose whether the individual has Alzheimer's Disease.

23. The method of claim 22, wherein the level of the biomarkers, age, gender, and ApoE genotype are incorporated into an algorithm to diagnose whether the individual has Alzheimer's Disease.

24. The method of claim 23, wherein said algorithm is selected from the group consisting of Logistic Regression, Boosted Tree Models, Flexible Discriminant Analysis (FDA), K-Nearest Neighbors (KNN), Naive Bayes, Partial Least Squares (PLS), Randon Forests, Shrunken Centroids, Sparse Partial Least Squares, Support Vector Machines approaches, and any combination thereof.

25. A method of determining risk of an individual developing Alzheimer's disease, the method comprising:

measuring the level of one or more biomarkers selected from Table 2 in a plasma sample of an individual, and comparing the level of the biomarkers in the plasma sample of the individual to positive and optionally to negative control values, wherein levels of the biomarkers in the plasma sample of the individual that resemble the positive control and do not resemble the negative control are an indication that said individual is at risk of developing Alzheimer's disease.

26. The method of claim 25, further comprising analyzing age, gender, and ApoE genotype (e2/e2, e2/e3, e2/e4, e3/e3, e3/e4 and e4/e4) of the individual in order to determine risk of an individual of developing Alzheimer's disease.

27. The method of claim 26, wherein the level of biomarkers, age, gender, and ApoE genotype are incorporated into an algorithm in order to determine risk of an individual of developing Alzheimer's disease.

28. The method of claim 27, wherein said algorithm is selected from the group consisting of Logistic Regression, Boosted Tree Models, Flexible Discriminant Analysis (FDA), K-Nearest Neighbors (KNN), Naive Bayes, Partial Least Squares (PLS), Randon Forests, Shrunken Centroids, Sparse Partial Least Squares, Support Vector Machines approaches, and any combination thereof.

29. A method for monitoring Alzheimer's disease in an individual, the method comprising:

measuring the level of one or more biomarkers selected from Table 2 in a plasma sample of an individual at a first time, and comparing the level of the biomarkers in the plasma sample of the individual at the first time to a level of the biomarkers in a plasma sample of the individual at the second time.

30. The method of claim 29, further comprising analyzing age, gender, and ApoE genotype (e2/e2, e2/e3, e2/e4, e3/e3, e3/e4 and e4/e4) of the individual in order to monitor Alzheimer's disease in the individual.

31. The method of claim 30, wherein the level of the biomarkers, age, gender, and ApoE genotype are incorporated into an algorithm in order to monitor Alzheimer's disease in the individual.

32. The method of claim 31, wherein said algorithm is selected from the group consisting of Logistic Regression, Boosted Tree Models, Flexible Discriminant Analysis (FDA), K-Nearest Neighbors (KNN), Naive Bayes, Partial Least Squares (PLS), Randon Forests, Shrunken Centroids, Sparse Partial Least Squares, Support Vector Machines approaches, and any combination thereof.

33. A method of assessing increased risk of developing Alzheimer's disease, said method comprising:

obtaining a first sample of plasma from an individual blood at a first time, assessing the level of a biomarker for Alzheimer's disease in said first plasma sample to obtain a baseline level,

obtaining a second sample of plasma from said individual at a second time to obtain a second level,

assessing the level of said biomarker in said second plasma sample to obtain a second level,

wherein if the second level is greater in the case of a biomarker that is over-expressed in AD or lower in the case of a biomarker that is under-expressed in AD, the individual is at an increased risk of developing Alzheimer's disease.

34. The method of claim 33, wherein the biomarker is one or more biomarkers selected from Table 2.

35. The method of claim 34, wherein age, gender, and ApoE genotype (e2/e2, e2/e3, e2/e4, e3/e3, e3/e4 and e4/e4) of the individual is analyzed in order to assess whether the individual is at an increased risk of developing Alzheimer's disease.

36. The method of claim 35 wherein the level of the biomarkers, age, gender, and ApoE genotype are incorporated into an algorithm in order to assess whether the individual is at an increased risk of developing Alzheimer's disease.

37. The method of claim 36, wherein said algorithm is selected from the group consisting of Logistic Regression, Boosted Tree Models, Flexible Discriminant Analysis (FDA), K-Nearest Neighbors (KNN), Naive Bayes, Partial Least Squares (PLS), Randon Forests, Shrunken Centroids, Sparse Partial Least Squares, Support Vector Machines approaches, and any combination thereof.

38. A method of assessing the likelihood that a pharmaceutical agent is efficacious in treating Alzheimer's disease in an individual, said method comprising:

obtaining a first sample of plasma from an individual in the absence of a pharmaceutical agent,

assessing the level of a biomarker for Alzheimer's disease in said first plasma sample to obtain a baseline level,

administering said pharmaceutical agent to said individual, obtaining a second sample of plasma from said individual after administration of said pharmaceutical agent,

assessing the level of said biomarker for Alzheimer's disease in said second plasma sample to obtain a treated level,

wherein the likelihood that the pharmaceutical agent treats Alzheimer's disease is increased if the treated level is lower than the baseline level in the case of a biomarker that is over-expressed in AD and higher than the baseline level in the case of a biomarker that is under-expressed in AD.

39. The method of claim 38, wherein the biomarker is one or more biomarkers selected from Table 2.

40. The method of claim 39, wherein age, gender, and ApoE genotype (e2/e2, e2/e3, e2/e4, e3/e3, e3/e4 and e4/e4) of the individual is analyzed in order to assess the likelihood that the pharmaceutical agent is efficacious in treating Alzheimer's disease in the individual.

41. The method of claim 40, wherein the levels of the biomarkers, age, gender, and ApoE genotype are incorporated into an algorithm in order to assess the likelihood that the pharmaceutical agent is efficacious in treating Alzheimer's disease in the individual.

42. The method of claim 41, wherein said algorithm is selected from the group consisting of Logistic Regression, Boosted Tree Models, Flexible Discriminant Analysis (FDA), K-Nearest Neighbors (KNN), Naive Bayes, Partial Least Squares

(PLS), Randon Forests, Shrunken Centroids, Sparse Partial Least Squares, Support Vector Machines approaches, and any combination thereof.

43. A method of diagnosing Alzheimer's Disease in an individual, comprising

obtaining a plasma sample from said individual;

assaying levels of at least five biomarkers in said plasma sample, wherein said at least five biomarkers comprise: Cortisol, IGF BP2, pancreatic polypeptide, resistin, and ApoE; and comparing said levels of each of said at least five biomarkers to non-AD controls, wherein a statistically significant increase in said levels of Cortisol, IGF BP2, pancreatic polypeptide, and resistin, and a statistically significant decrease in said level of ApoE provides a positive diagnosis of AD in said individual.

44. A method of diagnosing Alzheimer's Disease in an individual, comprising

obtaining a plasma sample from said individual;

assaying levels of at least five biomarkers in said plasma sample, wherein said at least five biomarkers comprise: Cortisol, IGF BP2, pancreatic polypeptide, resistin, and osteopontin; and

comparing said levels of each of said at least five biomarkers to non-AD controls, wherein a statistically significant increase in said levels of Cortisol, IGF BP2, pancreatic polypeptide, osteopontin and resistin provides a positive diagnosis of AD in said individual.

45. The method of claim 43 or 44, further comprising conducting one or more cognitive tests on said individual to confirm said positive diagnosis of AD.

46. The method of claim 43, 44 or 45, further comprising obtaining a CSF sample from said individual to confirm said positive diagnosis of AD.

47. A method for screening to identify individuals at increased risk of developing AD, the method comprising

obtaining a plasma sample from said individual;

assaying levels of at least five biomarkers in said plasma sample, wherein said at least five biomarkers comprise: Cortisol, IGF BP2, pancreatic polypeptide, resistin, and ApoE; and

comparing said levels of each of said at least five biomarkers to non-AD controls, wherein a statistically significant increase in said levels of Cortisol, IGF BP2, pancreatic polypeptide, and resistin, and a statistically significant decrease in said level of ApoE provides an identification of an individual at increased risk of developing AD.

48. A method for screening to identify individuals at increased risk of developing AD, the method comprising

obtaining a plasma sample from said individual;

comparing said levels of each of said at least five biomarkers to non-AD controls, wherein a statistically significant increase in said levels of Cortisol, IGF BP2, pancreatic polypeptide, osteopontin and resistin provides an identification of an individual at increased risk of developing AD.

49. A method of diagnosing Alzheimer's Disease in an individual, comprising

obtaining a plasma sample from said individual;

determining whether, relative to non-AD controls, said levels of said at least five biomarkers provide a signature of a positive diagnosis of AD in said individual, wherein said signature comprises: a statistically significant increase in said levels of each of Cortisol, IGF BP2, pancreatic polypeptide, and resistin, and a statistically significant decrease in said level of ApoE.

50. A method of diagnosing Alzheimer's Disease in an individual, comprising

obtaining a plasma sample from said individual;

determining whether, relative to non-AD controls, said levels of said at least five biomarkers provide a signature of a positive diagnosis of AD in said individual, wherein said signature comprises: a statistically significant increase in said levels of each of Cortisol, IGF BP2, pancreatic polypeptide, osteopontin and resistin.

51. A kit for detecting Alzheimer's Disease in an individual with at least 80% sensitivity, comprising a reagent for detecting a plurality of biomarkers for Alzheimer's Disease in a biological sample from the individual.

52. The kit of claim 51 , wherein said reagent is capable of detecting biomarkers of Alzheimer's disease in a plasma sample.

53. The kit of claim 51, wherein said biomarkers of Alzheimer's disease comprises Cortisol, pancreatic polypeptide, osteopontin, IGF BP2, and resistin.

54. The kit of claim 51, wherein said biomarkers for Alzheimer's disease comprise resistin, pancreatic polypeptide, ApoD, G-CSF, MlPlbeta, ApoE, NrCAM, MMP2, Cortisol, PDGF-BB, MMP1, and 1-309.

55. The kit of claim 51, wherein said biomarkers for Alzheimer's disease comprise Alpha- 1 Microglobulin, Angiopoietin-2, Apolipoprotein E, Beta-2

Microglobulin, Chemokine (C-X-C motif) ligand 13 (CXCL13) also known as B lymphocyte chemoattractant (BLC), Cortisol, E-Selectin, FAS, Fatty Acid Binding Protein, Hepatocyte Growth Factor, IGF BP-2, Interleukin 10, NT-Pro-BNP (Brain natriuretic peptide co-secreted along with a 76 amino acid N-terminal fragment),

Osteopontin, Pancreatic Polypeptide, PAPP-A, Resistin, Stem Cell Factor, Tenascin C, Thrombomodulin, tissue inhibitor of metalloproteinases-1 (TIMP-1), vascular cell adhesion molecule- 1 (VCAM-1), Vascular endothelial growth factor (VEGF), and Von Willebrand Factor.

56. The kit of claim 51, wherein said biomarkers is one or more biomarkers selected from Table 2.

57. The kit of claim 51, wherein the reagent is an antibody.

58. The kit of claim 51 or claim 57, wherein the reagent is on a solid support.

The kit of claim 51 or 57 comprising an instruction manual.