Methods for predicting cognitive decline in a subject

Abstract

Alzheimer's disease is strongly linked to biological aging and bioenergetic abnormalities. Systemic dysregulation of metabolism is a hallmark of the physiological decline of tissues with age. We aimed to explore untargeted metabolomic profiling of blood samples from amyloid-positive people to distinguish individuals who progressed to cognitive decline from those who remained cognitively intact despite having amyloid deposits in the brain. A minimal signature of 9 metabolites identified decliners and non-decliners of cognitive function in participants with an amyloid load. These findings are of clinical importance as they suggest that a metabolic fingerprint may help to predict patients who will develop cognitive decline. Due to the high prevalence of brain amyloid-positivity in older adults, identifying adults who will have cognitive decline will enable the development of personalized and early interventions. The present invention relates to an in vitro method for predicting cognitive decline in a subject comprising the step of determining the level of at least one metabolite selected in the group consisting of 3-hydroxybutyrate, acetone, triglyceride 48:3, glucose, citrate, succinate, methionine, serine, sphingomyelin d18:1 / C26:0 in a biological sample obtained from the subject.

Description

[0001] METHODS FOR PREDICTING COGNITIVE DECLINE IN A SUBJECT

[0002] FIELD OF THE INVENTION:

[0003] The present invention is in the field of medicine and relates in particular to cognition.

[0004] BACKGROUND OF THE INVENTION:

[0005] Alzheimer's disease (AD) affects over 35 million people worldwide, and is anticipated to affect 115 million people by 2050 (1). To date, no treatment can reverse clinical progression of the disease, especially in its later stages. Early detection is crucial from a clinical and societal point of view, potentially enabling the use of preventive strategies to fight against memory loss, cognitive decline, and functional impairment. Amyloid-beta (A0) peptide accumulation is thought to be an early trigger and marker of AD pathophysiology, and total amyloid load increases the risk of cognitive decline onset (2). However, only some individuals with amyloid accumulation experience cognitive decline. The identification of patients with an increased risk of cognitive decline is a major challenge to better determine those who might best benefit from innovative therapies (3) and will require additional metrics beyond amyloid load for better prediction.

[0006] AD is strongly linked to aging, which is accompanied by cognitive decline, memory loss, metabolic dysregulation, bioenergetic abnormalities, and inflammation. Studies have linked metabolomic profiles to aging (4), disease onset (5) (6), and mortality (7), demonstrating that the human blood metabolome directly reflects physiological status. However, although perturbations in metabolism are widely recognized to be related to aging, few links have been established between systemic abnormalities in metabolism and cognitive decline (8). The healthy brain is the key organ that controls an individual's homeostasis, and its energy metabolism is fueled exclusively by metabolites such as glucose and ketone bodies. Furthermore, the lipid rheostat is essential for proper brain function. Therefore, the global monitoring of both lipid and polar metabolites is essential to assess as early as possible the decline of brain health, which is usually manifested by cognitive decline. In this context, metabolomics has become a powerful phenotyping tool, in which measurements of metabolites at scale enables a molecular understanding of (patho)physiology and identification of biomarkers of metabolic deviations (9) (10) (11). Compared to clinical assessments based on single metabolites, metabolic signatures provide a direct input and readout of aging processes, which can reveal subtle key metabolic changes and ultimately stratify health trajectories (12). Several metabolomics studies using proton (1H) nuclear magnetic resonance (NMR) or mass spectrometry (MS) on blood samples have identified many altered metabolites in patients with dementia compared with healthy controls (13) (14). Nevertheless, these studies were cross- sectional, capturing single timepoints. Longitudinal studies considering development of cognitive decline that occurs over a long period of time could connect metabolic changes to clinical phenotypes.

[0007] SUMMARY OF THE INVENTION:

[0008] The present invention is defined by the claims. In particular, the present invention relates to an in vitro method for predicting cognitive decline in a subject comprising the step of determining the level of at least one metabolite selected in the group consisting of 3- hydroxybutyrate, acetone, triglyceride 48:3, glucose, citrate, succinate, methionine, serine, sphingomyelin dl8: l / C26:0 in a biological sample obtained from the subject.

[0009] DETAILED DESCRIPTION OF THE INVENTION:

[0010] In this study we identified a panel of metabolic biomarkers that could be used to predict cognitive decline at early stages of amyloid accumulation before clinical symptoms manifest. To fulfil this objective, amyloid positive subjects from the MAPT study, categorized as cognitive decliners or non-decliners over up to a 4-year follow-up, were selected and plasma samples at inclusion were analysed by determining the plasma-derived metabolite signature of each individual. We identified a minimum signature of polar and lipids metabolites that could both predict cognitive decline and provide information on the putative mechanisms leading to AD pathophysiology.

[0011] Methods for predicting cognitive decline

[0012] In a first aspect, the present invention relates to an in vitro method for predicting cognitive decline in a subject comprising the step of determining the level of at least one metabolite selected in the group consisting of 3 -hydroxybutyrate, acetone, triglyceride 48:3, glucose, citrate, succinate, methionine, serine, sphingomyelin dl8: l / C26:0 in a biological sample obtained from the subject. In some embodiments, the method further encompasses a step of comparing the levels of 3 -hydroxybutyrate, acetone, triglyceride 48:3, glucose, citrate, succinate, methionine, serine, sphingomyelin dl8: l / C26:0 with predetermined reference values. As used herein, the term “subject” refers to any mammals, such as a rodent, a feline, a canine or a primate. In a preferred embodiment, the subject is a human. In some embodiments, the subject is an adult aged forty-five or over. In some embodiments, the subject is an elderly subject. As used herein the term “elderly subject” refers to an adult aged sixty or over. In some embodiments, the subject suffers from attention, learning, intelligence, language, memory, perception, judgement, mental acuity, decision making, problem solving or reasoning trouble. In some embodiments, the subject suffers or is at risk of suffering from cognitive decline. In some embodiments, the subject suffers or has suffers from a neurodegenerative condition (e.g. dementia, Alzheimer’s disease, Parkinson’s disease), an infection (e.g. COVID-19, HIV, toxoplasmosis), a physical or psychological trauma (e.g. aggression, post-traumatic stress, depression) or has undergone a medical treatment (e.g. administration of an anesthetic, narcotics, surgery). In some embodiments, the subject suffers or is at risk of suffering from age- related cognitive decline. In some embodiments, the subject is characterized as cognitive decliner by at least one clinical test. In some embodiments, the subject is an amyloid-positive subject. As used herein, the term “amyloid-positive subject” refers to a subject having P- amyloid peptide accumulation in the brain. In some embodiments, the subject suffers or is at risk of suffering from Alzheimer’s disease.

[0013] As used herein, the term “cognition” has its general meaning in the art and refers to a set of mental processes involved in brain functions related to knowledge (e.g. attention, learning, intelligence, language, memory, perception, judgement, mental acuity, decision making, problem solving, reasoning...).

[0014] As used herein, the term “cognitive decline” refers to a reduction in one or more cognitive abilities. In order to measure cognitive decline, numerous clinical tests are available. As example, the Mini-Mental State Examination (MMSE®) or "Folstein test” is a test of cognitive evaluation and memory capacities which helps to detect the presence of dementia or to monitor cognitive evolution (« Mini-Mental State: A Practical Method for Grading the Cognitive State of Patients for the Clinician », Folstein M. et al., Journal of Psychiatric Research, 1975, vol. 12, n° 3, pp. 189-198). Other examples include the “Montreal Cognition Assessment” or “MOCA®”, the “General Practitioner Cognition” or “GP-COG”, the “6-item impairment test” or “6-CIT” or the “clock-drawing test”. Typically, the term “predicting cognitive decline” includes detection of cognitive decline before clinical symptoms, before it can be diagnosed by clinical tests or monitoring cognitive evolution to detect cognitive decline in order to reverse clinical progression of the cognitive decline.

[0015] As used herein, the term “biological sample” refers to any biological sample obtained from the subject for the purpose of evaluation in vitro. In some embodiments, the biological sample is a body fluid sample. Examples of body fluids are blood, serum, plasma, amniotic fluid, brain / spinal cord fluid, liquor, cerebrospinal fluid, sputum, throat and pharynx secretions and other mucous membrane secretions, synovial fluids, ascites, tear fluid, lymph fluid and urine. As used herein, the term “blood sample” means a whole blood sample obtained from the subject. In some embodiments, the biological sample is a blood sample, a plasma sample or a serum sample. In a preferred embodiment, the biological sample is a plasma sample. In the context of the invention, the term “a biological sample” may be understood as a single biological sample for every metabolites, one biological sample per metabolite when the level of two or more metabolites is determined or at least two biological samples when the level of two or more metabolites is determined.

[0016] In the context of the present invention, the term “at least one metabolite” encompasses one, two, three, four, five, six, seven, eight, nine or more metabolites. In some embodiments, at least one metabolite is selected in the group consisting of 3 -hydroxybutyrate, acetone, citrate, succinate, methionine and at least one further metabolite is selected in the group consisting of glucose, serine, triglyceride 48:3 or sphingomyelin dl8: l / C26:0.

[0017] In some embodiments, the present invention relates to an in vitro method for predicting cognitive decline in a subject comprising the step of determining the level of the metabolites selected in the group consisting of 3 -hydroxybutyrate, acetone, triglyceride 48:3, glucose, citrate, succinate, methionine, serine, sphingomyelin dl8:l / C26:0, in a biological sample obtained from the subject.

[0018] In some embodiments, the present invention relates to an in vitro method for predicting cognitive decline in a subject comprising the step of: i) determining the level of at least one metabolite selected in the group consisting of 3 -hydroxybutyrate, acetone, triglyceride 48:3, glucose, citrate, succinate, methionine, serine, sphingomyelin dl8: l / C26:0 in a biological sample obtained from the subject; ii) comparing the level determined at step i) with a predetermined reference value and iii) concluding that the subject is at risk of suffering from cognitive decline when the level of 3 -hydroxybutyrate, acetone, triglyceride 48:3, glucose, citrate, succinate, methionine, serine and / or sphingomyelin dl8: l / C26:0 determined at step i) is modulated (i.e. higher or lower) as compared to the predetermined reference value.

[0019] In some embodiments, the present invention relates to an in vitro method for predicting cognitive decline in a subject comprising the step of: i) determining the level of at least one metabolite selected in the group consisting of 3 -hydroxybutyrate, acetone, triglyceride 48:3, glucose, citrate, succinate, methionine, serine, sphingomyelin dl8: l / C26:0 in a biological sample obtained from the subject; ii) comparing the level determined at step i) with a predetermined reference value and iii) concluding that the subject is at risk of suffering from cognitive decline when the level of 3 -hydroxybutyrate, acetone, citrate, succinate and / or methionine determined at step i) is lower than the predetermined reference value and / or when the level of glucose, serine, triglyceride 48:3 and / or sphingomyelin dl8: l / C26:0 determined at step i) is higher than the predetermined reference value.

[0020] In some embodiments, the present invention relates to an in vitro method for predicting cognitive decline in a subject comprising the step of: i) determining the level of the metabolites selected in the group consisting of 3- hydroxybutyrate, acetone, triglyceride 48:3, glucose, citrate, succinate, methionine, serine, sphingomyelin dl8: l / C26:0 in a biological sample obtained from the subject; ii) comparing the levels determined at step i) with their predetermined reference values and iii) concluding that the subject is at risk of suffering from cognitive decline when the levels of 3 -hydroxybutyrate, acetone, citrate, succinate and methionine determined at step i) are lower than the predetermined reference values and when the levels of glucose, serine, triglyceride 48:3 and sphingomyelin dl8: l / C26:0 determined at step i) are higher than the predetermined reference values.

[0021] In some embodiments, the present invention relates to an in vitro method for predicting cognitive decline in a subject comprising the step of: i) determining the level of at least one metabolite selected in the group consisting of 3 -hydroxybutyrate, acetone, triglyceride 48:3, glucose, citrate, succinate, methionine, serine, sphingomyelin dl8: l / C26:0 in a first biological sample obtained from the subject; ii) comparing the level determined at step i) with the level determined in a second biological sample obtained from the subject; and iii) concluding that the subject is at risk of suffering from cognitive decline when the level of 3 -hydroxybutyrate, acetone, triglyceride 48:3, glucose, citrate, succinate, methionine, serine and / or sphingomyelin dl8: l / C26:0 determined at step ii) is modulated (i.e. higher or lower) as compared to the level determined at step i).

[0022] In some embodiments, the present invention relates to an in vitro method for predicting cognitive decline in a subject comprising the step of: i) determining the level of at least one metabolite selected in the group consisting of 3 -hydroxybutyrate, acetone, triglyceride 48:3, glucose, citrate, succinate, methionine, serine, sphingomyelin dl8: l / C26:0 in a first biological sample obtained from the subject; ii) comparing the level determined at step i) with the level determined in a second biological sample obtained from the subject; and iii) concluding that the subject is at risk of suffering from cognitive decline when the level of 3 -hydroxybutyrate, acetone, citrate, succinate and / or methionine determined at step ii) is lower than the level determined at step i) and / or when the level of glucose, serine, triglyceride 48:3 and / or sphingomyelin dl8: l / C26:0 determined at step ii) is higher than the level determined at step i).

[0023] In some embodiments, the present invention relates to an in vitro method for predicting cognitive decline in a subject comprising the step of: i) determining the level of the metabolites selected in the group consisting of 3- hydroxybutyrate, acetone, triglyceride 48:3, glucose, citrate, succinate, methionine, serine, sphingomyelin dl8: l / C26:0 in a first biological sample obtained from the subject; ii) comparing the levels determined at step i) with the levels determined in a second biological sample obtained from the subject; and iii) concluding that the subject is at risk of suffering from cognitive decline when the levels of 3 -hydroxybutyrate, acetone, citrate, succinate and methionine determined at step ii) are lower than the levels determined at step i) and when the levels of glucose, serine, triglyceride 48:3 and sphingomyelin dl8: l / C26:0 determined at step ii) are higher than the levels determined at step i).

[0024] In some embodiments, the present invention relates to an in vitro method for monitoring cognitive decline in a subject comprising the step of: i) determining the level of at least one metabolite selected in the group consisting of 3 -hydroxybutyrate, acetone, triglyceride 48:3, glucose, citrate, succinate, methionine, serine, sphingomyelin dl8: l / C26:0 in a biological sample obtained from the subject; ii) comparing the level determined at step i) with a predetermined reference value and iii) concluding that the subject is at risk of suffering from cognitive decline when the level of 3 -hydroxybutyrate, acetone, triglyceride 48:3, glucose, citrate, succinate, methionine, serine and / or sphingomyelin dl8: l / C26:0 determined at step i) is modulated (i.e. higher or lower) as compared to the predetermined reference value.

[0025] In some embodiments, the present invention relates to an in vitro method for monitoring cognitive decline in a subject comprising the step of: i) determining the level of at least one metabolite selected in the group consisting of 3 -hydroxybutyrate, acetone, triglyceride 48:3, glucose, citrate, succinate, methionine, serine, sphingomyelin dl8: l / C26:0 in a biological sample obtained from the subject; ii) comparing the level determined at step i) with a predetermined reference value and iii) concluding that the subject is at risk of suffering from cognitive decline when the level of 3 -hydroxybutyrate, acetone, citrate, succinate and / or methionine determined at step i) is lower than the predetermined reference value and / or when the level of glucose, serine, triglyceride 48:3 and / or sphingomyelin dl8: l / C26:0 determined at step i) is higher than the predetermined reference value.

[0026] In some embodiments, the present invention relates to an in vitro method for monitoring cognitive decline in a subject comprising the step of: i) determining the level of the metabolites selected in the group consisting of 3- hydroxybutyrate, acetone, triglyceride 48:3, glucose, citrate, succinate, methionine, serine, sphingomyelin dl8: l / C26:0 in a biological sample obtained from the subject; ii) comparing the levels determined at step i) with their predetermined reference values and iii) concluding that the subject is at risk of suffering from cognitive decline when the levels of 3 -hydroxybutyrate, acetone, citrate, succinate and methionine determined at step i) are lower than the predetermined reference values and when the levels of glucose, serine, triglyceride 48:3 and sphingomyelin dl8: l / C26:0 determined at step i) are higher than the predetermined reference values.

[0027] In some embodiments, the present invention relates to an in vitro method for monitoring cognitive decline in a subject comprising the step of: i) determining the level of at least one metabolite selected in the group consisting of 3 -hydroxybutyrate, acetone, triglyceride 48:3, glucose, citrate, succinate, methionine, serine, sphingomyelin dl8: l / C26:0 in a first biological sample obtained from the subject; ii) comparing the level determined at step i) with the level determined in a second biological sample obtained from the subject and iii) concluding that the subject is at risk of suffering from cognitive decline when the level of 3 -hydroxybutyrate, acetone, triglyceride 48:3, glucose, citrate, succinate, methionine, serine and / or sphingomyelin dl8: l / C26:0 determined at step ii) is modulated (i.e. higher or lower) as compared to the level determined at step i). In some embodiments, the present invention relates to an in vitro method for monitoring cognitive decline in a subject comprising the step of: i) determining the level of at least one metabolite selected in the group consisting of 3 -hydroxybutyrate, acetone, triglyceride 48:3, glucose, citrate, succinate, methionine, serine, sphingomyelin dl8: l / C26:0 in a first biological sample obtained from the subject; ii) comparing the level determined at step i) with the level determined in a second biological sample obtained from the subject and iii) concluding that the subject is at risk of suffering from cognitive decline when the level of 3 -hydroxybutyrate, acetone, citrate, succinate and / or methionine determined at step ii) is lower than the level determined at step i) and / or when the level of glucose, serine, triglyceride 48:3 and / or sphingomyelin dl8: l / C26:0 determined at step ii) is higher than the level determined at step i).

[0028] In some embodiments, the present invention relates to an in vitro method for monitoring cognitive decline in a subject comprising the step of: i) determining the level of the metabolites selected in the group consisting of 3- hydroxybutyrate, acetone, triglyceride 48:3, glucose, citrate, succinate, methionine, serine, sphingomyelin dl8: l / C26:0 in a first biological sample obtained from the subject; ii) comparing the levels determined at step i) with the levels determined in a second biological sample obtained from the subject and iii) concluding that the subject is at risk of suffering from cognitive decline when the levels of 3 -hydroxybutyrate, acetone, citrate, succinate and methionine determined at step ii) are lower than the levels determined at step i) and when the levels of glucose, serine, triglyceride 48:3 and sphingomyelin dl8: l / C26:0 determined at step ii) are higher than the levels determined at step i).

[0029] As example, the present method for monitoring cognitive decline in a subject could be particularly useful to monitor cognitive decline in a subject amyloid-positive with no clinical symptoms of cognitive decline, e.g. by determining the level of the metabolites every 3 years. In another example, the present method for monitoring cognitive decline in a subject could be particularly useful in a young subject, e.g. by determining the level of the metabolites every 10 years. In another example, the present method for monitoring cognitive decline in a subject could be particularly useful in a subject aged sixty-five and over, e.g. by determining the level of the metabolites every years.

[0030] In some embodiments, the second sample is obtained from the subject at least 1, 2, 3, 4 or 5 weeks after the first one. In some embodiments, the second sample is obtained from the subject at least 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11 or 12 months after the first one. In some embodiments, the second sample is obtained from the subject at least 1, 2, 3, 4, 5, 6, 7, 8, 9 or 10 years after the first one.

[0031] As used herein, the term “predetermined reference value” refers to a threshold value or a cut-off value. A "threshold value", “reference value” or "cut-off value" can be determined experimentally, empirically, or theoretically. A threshold value can also be arbitrarily selected based upon the existing experimental and / or clinical conditions, as would be recognized by a person of ordinary skilled in the art. For example, retrospective measurement of the level of the marker of the invention (e.g. the at least one metabolite) in properly banked historical patient samples may be used in establishing the predetermined corresponding reference value. In some embodiments, the predetermined corresponding reference value is the median measured in the population of the patients for the marker of in the invention. In some embodiments, the threshold value has to be determined in order to obtain the optimal sensitivity and specificity according to the function of the test and the benefit / risk balance (clinical consequences of false positive and false negative). Typically, the optimal sensitivity and specificity (and so the threshold value) can be determined using a Receiver Operating Characteristic (ROC) curve based on experimental data. For example, after determining the level of the marker of the invention in a group of reference, one can use algorithmic analysis for the statistic treatment of the levels determined in samples to be tested, and thus obtain a classification standard having significance for sample classification. The full name of ROC curve is receiver operator characteristic curve, which is also known as receiver operation characteristic curve. It is mainly used for clinical biochemical diagnostic tests. ROC curve is a comprehensive indicator that reflects the continuous variables of true positive rate (sensitivity) and false positive rate (1- specificity). It reveals the relationship between sensitivity and specificity with the image composition method. A series of different cut-off values (thresholds or critical values, boundary values between normal and abnormal results of diagnostic test) are set as continuous variables to calculate a series of sensitivity and specificity values. Then sensitivity is used as the vertical coordinate and specificity is used as the horizontal coordinate to draw a curve. The higher the area under the curve (AUC), the higher the accuracy of diagnosis. On the ROC curve, the point closest to the far upper left of the coordinate diagram is a critical point having both high sensitivity and high specificity values. The AUC value of the ROC curve is between 1.0 and 0.5. When AUC>0.5, the diagnostic result gets better and better as AUC approaches 1. When AUC is between 0.5 and 0.7, the accuracy is low. When AUC is between 0.7 and 0.9, the accuracy is moderate. When AUC is higher than 0.9, the accuracy is quite high. This algorithmic method is preferably done with a computer. Existing software or systems in the art may be used for the drawing of the ROC curve, such as: MedCalc 9.2.0.1 medical statistical software, SPSS 9.0, ROCPOWER.SAS, DESIGNROC.FOR, MULTIREADER POWER. SAS, CREATE-ROC.SAS, GB STAT VIO.O (Dynamic Microsystems, Inc. Silver Spring, Md., USA), etc.

[0032] In some embodiments, the level of at least one metabolite is determined by mass spectrometry. Mass spectrometry (MS) is an analytical technique used to measure a mass-to-charge ratio of ions in pure samples as well as complex mixtures. Results are depicted in a spectrum. The spectra are used to determine the elemental or isotopic signature of a sample, the masses of particles and of molecules, and to elucidate the chemical identity or structure of molecules and other chemical compounds. In a typical MS procedure, the sample may be solid, liquid, or gaseous. The sample is ionized, for example by bombarding it with a beam of electrons. Some of the sample's molecules break up into positively charged fragments or simply become positively charged without fragmenting. These ions (fragments) are then separated according to their mass-to-charge ratio, for example by accelerating them and subjecting them to an electric or magnetic field. The ions are detected by a mechanism capable of detecting charged particles, such as an electron multiplier. Results are displayed as spectra of the signal intensity of detected ions as a function of the mass-to-charge ratio. The atoms or molecules in the sample can be identified by correlating known masses to the identified masses or through a characteristic fragmentation pattern. In some embodiments, the level of at least one metabolite is determined by high resolution mass spectrometry (HRMS). In some embodiments, the level of at least one metabolite is determined by high-performance liquid chromatography coupled with mass spectrometry (HPLC-MS). In some embodiments, the level of at least one metabolite is determined by supercritical fluid chromatography coupled with high-resolution mass spectrometry (SFC-HRMS). Supercritical fluid chromatography is typically used for the analysis and purification of low to moderate molecular weight. It is a normal phase chromatography that uses a supercritical fluid (e.g. carbon dioxide) as the mobile phase.

[0033] Cosolvents can be added to adapt the mobile phase polarity, such as example methanol.

[0034] In some embodiments, the level of at least one metabolite is determined by Nuclear Magnetic Resonance (NMR). NMR is a physical phenomenon wherein nuclei is in a strong constant magnetic field and is perturbated by a weak oscillating magnetic field. The nuclei respond by producing an electromagnetic signal with a frequency characteristic of the magnetic field at the nucleus. This process occurs when the oscillation frequency matches the intrinsic frequency of the nuclei which depends on the strength of the static magnetic field, the chemical environment, and the magnetic properties of the isotope involved. The most commonly used nuclei areJH and13C but19F,31P or33S can be studied by high field NMR spectroscopy as well.

[0035] Methods for treating cognitive decline

[0036] In order to reverse clinical progression of the cognitive decline, any of the methods of the present invention can comprise a further step of administering said subject with a treatment against cognitive decline or practising a non-drug treatment against cognitive decline.

[0037] Thus in another aspect, the present invention relates to a method of treating cognitive decline in a subject in need thereof comprising a step of determining the level of at least one metabolite selected in the group consisting of 3 -hydroxybutyrate, acetone, triglyceride 48:3, glucose, citrate, succinate, methionine, serine, sphingomyelin dl8: l / C26:0 in a biological sample obtained from the subject.

[0038] In some embodiments, the present invention relates to a method of treating cognitive decline in a subject in need thereof comprising the steps of: i) determining the level of at least one metabolite selected in the group consisting of 3 -hydroxybutyrate, acetone, triglyceride 48:3, glucose, citrate, succinate, methionine, serine, sphingomyelin dl8: l / C26:0 in a biological sample obtained from the subject; ii) comparing the level determined at step i) with a predetermined reference value; iii) concluding that the subject is at risk of suffering from cognitive decline when the level of 3 -hydroxybutyrate, acetone, triglyceride 48:3, glucose, citrate, succinate, methionine, serine and / or sphingomyelin dl8: l / C26:0 determined at step i) is modulated (i.e. higher or lower) as compared to the predetermined reference value; and iv) administering said subject with a treatment against cognitive decline or practising a non-drug treatment against cognitive decline.

[0039] In some embodiments, the present invention relates to a method of treating cognitive decline in a subject in need thereof comprising the steps of: i) determining the level of at least one metabolite selected in the group consisting of 3 -hydroxybutyrate, acetone, triglyceride 48:3, glucose, citrate, succinate, methionine, serine, sphingomyelin dl8: l / C26:0 in a biological sample obtained from the subject; ii) comparing the level determined at step i) with a predetermined reference value; iii) concluding that the subject is at risk of suffering from cognitive decline when the level of 3 -hydroxybutyrate, acetone, citrate, succinate and / or methionine determined at step i) is lower than the predetermined reference value and / or when the level of glucose, serine, triglyceride 48:3 and / or sphingomyelin dl8: l / C26:0 determined at step i) is higher than the predetermined reference value; and iv) Administering said subject with a treatment against cognitive decline or practising a non-drug treatment against cognitive decline.

[0040] In some embodiments, the present invention relates to a method of treating cognitive decline in a subject in need thereof comprising the steps of: i) determining the level of the metabolites selected in the group consisting of 3- hydroxybutyrate, acetone, triglyceride 48:3, glucose, citrate, succinate, methionine, serine, sphingomyelin dl8: l / C26:0 in a biological sample obtained from the subject; ii) comparing the levels determined at step i) with their predetermined reference values; iii) concluding that the subject is at risk of suffering from cognitive decline when the level of 3 -hydroxybutyrate, acetone, citrate, succinate and methionine determined at step i) are lower than the predetermined reference values and when the level of glucose, serine, triglyceride 48:3 and sphingomyelin dl8: l / C26:0 determined at step i) are higher than the predetermined reference values; and iv) administering said subject with a treatment against cognitive decline or practising a non-drug treatment against cognitive decline.

[0041] In some embodiments, the present invention relates to a method of treating cognitive decline in a subject in need thereof comprising the steps of: i) determining the level of at least one metabolite selected in the group consisting of 3 -hydroxybutyrate, acetone, triglyceride 48:3, glucose, citrate, succinate, methionine, serine, sphingomyelin dl8: l / C26:0 in a biological sample obtained from the subject; ii) comparing the level determined at step i) with the level determined in a second biological sample obtained from the subject; iii) concluding that the subject is at risk of suffering from cognitive decline when the level of 3 -hydroxybutyrate, acetone, triglyceride 48:3, glucose, citrate, succinate, methionine, serine and / or sphingomyelin dl8: l / C26:0 determined at step ii) is modulated (i.e. higher or lower) as compared to the level determined at step i) ; and iv) Administering said subject with a treatment against cognitive decline or practising a non-drug treatment against cognitive decline.

[0042] In some embodiments, the present invention relates to a method of treating cognitive decline in a subject in need thereof comprising the steps of: i) determining the level of at least one metabolite selected in the group consisting of 3 -hydroxybutyrate, acetone, triglyceride 48:3, glucose, citrate, succinate, methionine, serine, sphingomyelin dl8: l / C26:0 in a biological sample obtained from the subject; ii) comparing the level determined at step i) with the level determined in a second biological sample obtained from the subject; iii) concluding that the subject is at risk of suffering from cognitive decline when the level of 3 -hydroxybutyrate, acetone, citrate, succinate and / or methionine determined at step ii) is lower than the level determined at step i) and / or when the level of glucose, serine, triglyceride 48:3 and / or sphingomyelin dl8: l / C26:0 determined at step ii) is higher than the level determined at step i) ; and iv) Administering said subject with a treatment against cognitive decline or practising a non-drug treatment against cognitive decline. In some embodiments, the present invention relates to a method of treating cognitive decline in a subject in need thereof comprising the steps of: i) determining the level of the metabolites selected in the group consisting of 3- hydroxybutyrate, acetone, triglyceride 48:3, glucose, citrate, succinate, methionine, serine, sphingomyelin dl8: l / C26:0 in a biological sample obtained from the subject; ii) comparing the levels determined at step i) with the levels determined in a second biological sample obtained from the subject; iii) concluding that the subject is at risk of suffering from cognitive decline when the levels of 3 -hydroxybutyrate, acetone, citrate, succinate and methionine determined at step ii) are lower than the levels determined at step i) and when the levels of glucose, serine, triglyceride 48:3 and sphingomyelin dl8: l / C26:0 determined at step ii) are higher than the level determined at step i) ; and iv) Administering said subject with a treatment against cognitive decline or practising a non-drug treatment against cognitive decline.

[0043] Typically, treatments against cognitive decline are well-known and includes drug administration (e.g. cholinesterase inhibitors, donepezil, galantamine, rivastigmine tartrate, memantine hydrochloride, nonsteroidal anti-inflammatory drugs, vitamin E, statins) or nondrug treatment (e.g. functional rehabilitation, orthophonic or mortician therapy, behavioural therapy, cognitive stimulation, memory stimulation, sport, diets). In some embodiments, the treatment against cognitive decline is Lecanemab (BAN2401, Eisai). In some embodiments, the subject is amyloid positive and the treatment against cognitive decline is Lecanemab (BAN2401, Eisai).

[0044] As used herein, the terms “treating”, “treatment” or “therapy” refer to both prophylactic or preventive treatment as well as curative or disease modifying treatment, including treatment of subject at risk of contracting the disorder or suspected to have contracted the disorder as well as subject who are ill or have been diagnosed as suffering from a disease or medical condition, and includes suppression of clinical relapse. The term encompasses both drug administration and non-drug treatment. The treatment may be administered to a subject having a medical disorder or who ultimately may acquire the disorder, in order to prevent, cure, delay the onset of, reduce the severity of, or ameliorate one or more symptoms of a disorder or recurring disorder, or in order to prolong the survival of a subject beyond that expected in the absence of such treatment. By "therapeutic regimen" is meant the pattern of treatment of an illness, e.g., the pattern of dosing used during therapy. A therapeutic regimen may include an induction regimen and a maintenance regimen. The phrase "induction regimen" or "induction period" refers to a therapeutic regimen (or the portion of a therapeutic regimen) that is used for the initial treatment of a disease. The general goal of an induction regimen is to provide a high level of drug to a subject during the initial period of a treatment regimen. An induction regimen may employ (in part or in whole) a "loading regimen", which may include administering a greater dose of the drug than a physician would employ during a maintenance regimen, administering a drug more frequently than a physician would administer the drug during a maintenance regimen, or both. The phrase "maintenance regimen" or "maintenance period" refers to a therapeutic regimen (or the portion of a therapeutic regimen) that is used for the maintenance of a subject during treatment of an illness, e.g., to keep the subject in remission for long periods of time (months or years). A maintenance regimen may employ continuous therapy (e.g., administering a drug at regular intervals, e.g., weekly, monthly, yearly, etc.) or intermittent therapy (e.g., interrupted treatment, intermittent treatment, treatment at relapse, or treatment upon achievement of a particular predetermined criteria [e.g., disease manifestation, etc.]).

[0045] As used herein the terms "administering" or "administration" refer to the act of injecting or otherwise physically delivering a substance as it exists outside the body (e.g., treatment against cognitive decline) into the subject, such as by mucosal, intradermal, intravenous, subcutaneous, intramuscular delivery and / or any other method of physical delivery described herein or known in the art. When a disease, or a symptom thereof, is being treated, administration of the substance typically occurs after the onset of the disease or symptoms thereof. When a disease or symptoms thereof, are being prevented, administration of the substance typically occurs before the onset of the disease or symptoms thereof.

[0046] The invention will be further illustrated by the following figures and examples. However, these examples and figures should not be interpreted in any way as limiting the scope of the present invention.

[0047] FIGURES:

[0048] Figure 1: Unbiased plasma metabolomic profiling of either aqueous or lipid metabolites is not sufficient to predict cognitive decline. Two-dimensional PCA score plot of plasma samples (control, n = 12; decline, n = 8). (A) Integrated 'H NMR spectra of aqueous extracts (2 principal components; R2X = 55.6%). (B) SFC-HRMS spectra of lipidic extracts (2 principal components; R2X = 56.7%).

[0049] Figure !: 9 metabolites across the NMR and MS datasets together predict cognitive decline. Variable importance plots for the cognitive decline metabolome and lipidome biomarker panel. The most important variables (according to the absolute value of their coefficients) are ordered. (A) 1H-NMR: 3OHbut: 3 -hydroxybutyrate; (B) SFC-HRMS: TG: Triglyceride; SM: Sphingomyelin.

[0050] Figure 3: Decline status can be predicted from a minimal metabolite signature. Score plot from multiblock sPLS-DA. Samples are represented based on the specified component (here 1 latent component). Samples are colored by decline status.

[0051] Figure 4: The multiblock sPLS-DA model accurately predicts groups without decline from groups with decline. Performance of the multiblock sPLS-DA model and Clustered Image Map (CIM) of metabo-clinical signatures. (A); ROC based on multiblock sPLS-DA model for the NMR dataset (AUC = 0.9167; p-value = 0.002028); (B) ROC based on multiblock sPLS-DA model for the Lipidomic dataset (AUC = 0.8021; p-value = 0. 02526); (C): Clustered Image Map for the variables selected by multiblock sPLS-DA on component 1. The CIM represents samples in rows (indicated by their Decline status on the left hand side of the plot) and selected features in columns (indicated by their data type at the top of the plot): 2.4237: 3 -hydroxybutyrate; 2.5126: Citrate; 2.3934: 3 -hydroxybutyrate; 2.4067: Succinate; 2.2399: Acetone; 2.6590: Methionine; 2.3253: 3 -hydroxybutyrate; 1.2039: 3 -hydroxybutyrate; 3.7474: Glucose; dl8.1_C26.0: Sphingomyelin; 4.0095: Serine; 48.3: Triglyceride.

[0052] EXAMPLE:

[0053] Material and Methods

[0054] Clinical Assessment:

[0055] MAPT Study

[0056] Plasmas samples were obtained from the Multidomain Alzheimer Preventive Trial (MAPT, ClinicalTrials.gov [NCT00672685]), a randomized, multicenter, placebo-controlled trial conducted with community-dwelling older adults in France and Monaco. Participants were allocated into 4 groups, either receiving co-3 polyunsaturated fatty acid (PUFA) supplementation, a multidomain intervention (based on cognitive training, nutritional counseling, and physical activity advice), both, or placebo. The intervention lasted for 3 years and was followed by an additional 2-year observational phase. Recruitment of participants started in May 2008 and ended in February 2011. Follow-up ended in April 2016 (15). Detailed description of the MAPT study can be found elsewhere (15). In summary, eligibility criteria comprised: age 70 years or older; not presenting major neurocognitive disorders, Mini -Mental State Examination [MMSE] score >24; presenting at least 1 of the following: spontaneous memory concern, inability to perform 1 instrumental activity of daily living (IADL), or slow usual-pace walking speed (<0.8 m / sec). The present study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guideline (16).

[0057] Ethical aspects

[0058] The study was approved by the French Ethics Committee located in Toulouse (CPP SOOM II) and authorized by the French Health Authority. Written consent was obtained from all participants. The protocol is registered on the clinical trials database (www.clinicaltrials.gov - NCT00672685).

[0059] Definition of sub-groups analysis

[0060] 24 sex-matched participants, half decliners, half non-decliners provided blood samples for this study, but only 20 participants had usable metabolomics data. Decliners / non-decliners were classified according to overtime decrease in the MMSE across the follow-up: those with > 2 points decrease were considered decliners; the others were considered non-decliners. The positive status of amyloid in plasma was measured as described previously (17).

[0061] Characteristics of the participants

[0062] Outcomes were assessed at the visit. Blood samples were collected at the initial visit and then annually for 5 years. Overall cognitive performance was assessed using: a composite cognitive score (CCS) (18) based on four tests (the 10 orientation items of the MMSE, the Digit Symbol Substitution Test (DSST), free and total recall of the Free and Cued Selective Reminding Test (FCSRT), and the Category Naming Test); and the MMSE score (19). CCS was computed as the mean z-score of the four domains, calculated using the baseline mean and SD values of corresponding test. The physical capacities were evaluated by the usual pace gait speed test (GS) , the Short Physical Performance Battery, the 5-repetition sit-to-stand test (5- STS) (20), and the maximal handgrip strength (HS) (measured in kg by a handled dynamometer [Jamar, Bolingbrook, IL]). Metabolomic profiling:

[0063] Extraction of plasma samples

[0064] 100 pl of plasma samples were homogenized in 1 ml of methanol. The homogenates were transferred to glass tubes with 900 pl of methanol and 2 ml of dichloromethane was added. The samples were vortexed and centrifuged (1500 g, 5 min, 4°C). Supernatants were collected and 600 pl of 0.9% NaCl solution was added. Samples were vortexed, centrifuged (1500 g, 5 min, 4°C), and aqueous and organic phases were collected into two test tubes.

[0065] Sample preparation of aqueous extracts for NMR analyses

[0066] Aqueous phases were dried using the SpeedVac facility. Dried extracts were reconstituted in 200 pl of phosphate buffer (0.2M, pH 7.0) prepared in deuterium oxide (D2O) and containing 1 mM of sodium trimethylsilylpropionate (TSP), centrifuged (15 min, 2870 g, 4°C) and transferred into 3 mm NMR tubes.

[0067] NMR analyses

[0068] 'H NMR spectra were obtained at 300 K on a Bruker Avance III HD 600 MHz NMR spectrometer (Bruker Biospin, Rheinstetten, Germany), operating at 600.13 MHz for 'H resonance frequency using an inverse detection 5 mmJH-13C-15N-3 IP cryoprobe. “Tuning” and “matching” of the probe, lock, shims tuning, pulse (90°) and gain computation are automatically performed on each sample. The 'H NMR spectra were acquired using the ID NOESY experiment with presaturation for water removal (noesyprld), with a mixing time of 100 ms. A total of 128 transients were collected into 64,000 data points using a spectral width of 12 ppm, a relaxation delay of 15 s and an acquisition time of 4.55 s. Prior to Fourier transform, an exponential line broadening function of 0.3 Hz was applied to the free induction decays. All NMR spectra were phase- and baseline-corrected and referenced to the chemical shift of TSP (0 ppm) using Topspin (V3.2, Bruker, Biospin, Germany). The 'H NMR spectra were then divided into variable size buckets between 8.5 and 0.7 ppm using the AMIX software (v3.9.15, Bruker, Rheinstetten, Germany), and area under the curve was calculated for each bucket (integration). Variable sized bucketing means that each bucket may have an individual size. Buckets were defined graphically as a spectral pattern (excluding solvent signals, and noise), and this pattern was used for bucketing. A total of 102 buckets or variables (several variables can correspond to the same metabolite) were defined with this method. Integrations were normalized according to the total intensity. Preprocessed data were then exported into an Excel file. Spectral assignment was based on matching one-dimensional (ID) data to reference spectra in a home-made database, as well as with other databases (https: / / bmrb.io / ; https: / / www.hmdb.ca; https: / / peakforest.org / ). Assignments were confirmed by two- dimensional (2D) NMR experiments:JH-13C HSQC (Heteronuclear Single Quantum Correlation);JH-JH COSY (Correlation Spectroscopy;JH-JH TOCSY (Total Correlation Spectroscopy) andJH-13C HMBC (Heteronuclear Multiple Bond Correlation). Validation of identification was based on the nomenclature of the metabolomics standards initiative (21). 30 compounds were identified (level 1), with the same proton and carbon- 13 chemical shifts of reference compounds analyzed in the same conditions.

[0069] Sample preparation for lipidomic analyses

[0070] 20% of the organic phase were dried off in the presence of a mixture of internal standards (11 :0 LPC (2.5mg / ml), 13:0 LPE (2mg / ml), 12:0 PG (2mg / ml), 13:0 PC( (5mg / ml), 12:0 PE (5.5mg / ml), dl 8: 1 / 12:0 Cer (2mg / ml), dl 8 : 1 / 12:0 lacCer (6mg / ml), dl 8 : 1 / 12:0 GalCer (2.5mg / ml), dl8:0 / 12:0 SM (5mg / ml), TG 17: 1 :17:0 / 17:0 d( (2.5mg / ml), FA 17:0 (5mg / ml), CE 17:0 (Img / ml), DG 12:0-12:0 (5mg / ml), PI 15:0-18:l-d7 (1.5mg / ml), PS 12:0-12:0 (3mg / ml)) and dissolved in lOOpl of MeOH:Isoprpanol:H2O (v / v / ; 65:35:5).

[0071] Untargeted lipidomic profiling by Supercritical Fluid chromatography coupled with high-resolution mass spectrometry (SFC-HRMS)

[0072] The lipid extract was profiled using a supercritical fluid chromatography. I L of the extract was injected on the Ultra-Performance Convergence Chromatography (UPC2) system coupled on-line to an Xevo G2-XS time of flight (Qtof) (Waters, Milford, MA) equipped with electrospray ionization (ESI). The analysis was performed in both ionization modes (positive and negative) in two separate runs on an ACQUITY UPC2 Torus diethylamine column (100 x 3.0 mm inner diameter (i.d.), particle size: sub-1.7 pm, Waters) at 40°C. Mobiles phases with a flow rate of 1.2 mL / min were constituted by SCCO2 for the A phase and Me0H:H20 (98:2; v / v) with 20 mM of ammonium acetate for the B phase (modifier). The gradient program was as follows: initial conditions were 1% of B solvent; from 0.5 min to 6 min it was increased to 40% then from 6 min to 6.10 min to 65%. The solvent B was maintained to 65 % during three minutes then the gradient went back to initial conditions in three minutes with an active back pressure regulator (AB PR), 1.500 pounds per square inch (psi). From 6 to 9 min, the flow rate was decreased to 1.1 mL / min. The make-up solvent was Me0H:H20 (95:5 ; v / v) at 0.1 mL / min during all run. The source parameters of the mass detector were set as follows: for positive and negative analysis source temperature was 150°C, capillary voltage was at -2.6 kV in negative mode and 3kV in positive mode, desolvation gas flow rate was 1000 L / Hr, cone gas flow rate was set at 50 L / Hr, and the desolvation temperature was 550°C. The analyses were performed in MS full scan in centroid mode from 50 to 1500 Da with dynamic range extended (DRE) activated. MS / MS experiments were performed in positive and negative ion modes on the same instrument, using a ramp of collision energy ranging from 10 to 50 eV. The isolation width was set at m / z 5. MS / MS mass spectra were inspected manually to confirm annotations. The method allows the separation and profiling of 18 subfamilies of lipids: sterol and sterol ester, diacylglycerid, triacylglycerid, ceramides, phosphatidylcholine (PC), phosphatidylserine, phosphatidylethanolamine (PE), phosphatidylinositol, phosphatidylglycerol, LysoPC, LysoPE, cardiolipin, sphingomyelin, monohexosylceramide, dihexosylceramides, free fatty acids and acyl-carnitine. The lipidomics data were processed with a suspect screening approach, through the interrogation of an in-house data base, using MS-DIAL (22) which allows the relative quantification of each species of sub classes families using internal standards (one standard per sub-class of lipids) (23).

[0073] Statistical analysis:

[0074] Clinical data

[0075] Descriptive statistics were provided using mean (SD) and absolute values and percentages, as appropriate. The Mann-Whitney U test and Fisher’s exact test were performed as appropriate to examine the differences between decliners in cognitive function and nondecliners.

[0076] Statistical Metabolonuc analysis

[0077] Multivariate analyses were used to separate patients according to decline status from metabolome or lipidome profiles. Firstly, Principal Components Analysis (PCA) was performed to reveal intrinsic clusters (for example, Sex) and detect eventual outliers. Partial least squares-discriminant analysis (PLS-DA) was then used to model the relationship between decline status and spectral data. Data were Pareto-scaled. The R2Y parameter represents the explained variance. Seven-fold cross validation was used to determine the number of latent variables to include in the PLS-DA model and to estimate the predictive ability (or predicted variance, Q2 parameter) of the fitted model. PLS-DA models with Q2 value higher than 0.4 were considered valid (24). SIMCA-P software (V14, Umetrics AB, Umea, Sweden) was used to perform the multivariate analyses and R (https: / / www.r-project.org / ) for univariate testing. The two analytical methods used to profile plasma metabolites provide complementary information, sinceJH NMR is used to profile polar metabolites in aqueous extracts and while SFC-HRMS allows to profiling lipids in organic extracts. Statistical integration of both datasets, i.e. simultaneous analysis of NMR and MS data sets, can be very beneficial to increase information about the decline status and so to get more predictive models of decline. We used DIABLO (25) for this purpose. This supervised multivariate method generalizes sparse Canonical Correlation Analysis (CCA) to classification. CCA is used to assess correlation between block variables, i.e. the NMR and MS blocks in this study. The sparse version of CCA performs variable selection. This selection allows discarding of noisy variables and redundancy within and between the data sets: only predictive variables of the decline status are selected in the final model. DIABLO works in two steps: in the first step, the optimal number of latent components is chosen. Then, the optimal number of variables to select in each dataset is fixed. These optimal numbers are chosen to minimize the balanced error rate (BER) defined as the average of the errors made on each class (Decliners classified as Non Decliners and Non Decliners classified as Decliners). BER was computed using 4-fold cross validation. We applied a bootstrap resampling strategy to assess stability of the optimal numbers of latent components and of variables to select. For each bootstrap sample, the optimal number of latent components to include in the model was firstly fixed (models with 1 to 5 components were tested). Once the number of latent components was fixed, the number of variables to select in each data set was optimized using a grid of values including 1 to 9 (step = 1), 10 to 100 (step = 5) variables for the metabolomic dataset and 1 to 9 (step = 1), 10 to 290 (step = 5) variables for the lipidomic dataset. The final model (on the entire sample) was fitted based on the bootstrap results. Model performance was evaluated using AUC (Aurea under Curve) values. The mixOmics R package was used for the DIABLO method (26).

[0078] Results

[0079] Characterization of the samples from MAPT study participants

[0080] 20 participants (n=8 decliners in cognitive function; n=12 non-decliners) were included from the placebo group of the MAPT study (Table 1). Decliners and non-decliners did not have baseline difference in socio-demographic factors, clinical measures, such as comorbidities, and functional measures that included gait speed, SPPB (Short Physical Performance Battery), mood, or cognitive function. Aqueous metabolomics or lipidomic profiling alone cannot predict future cognitive decline

[0081] We used dual technology for the same blood sample, which allows us to cover a very large number of metabolites and to assess the different lipid species. 'H-NMR provides relative quantification of metabolites based on the intensity of the spectral peaks while SFC-HRMS offers large metabolite coverage, sensitivity, and selectivity. So, to provide complementary information about aqueous and lipid metabolites and to broaden metabolite coverage, we performed unbiased plasma metabolomic profiling of aqueous and lipids metabolites, which allowed the identification of 30 polar metabolites and 290 lipid species. Individual PC A was first performed on each dataset to describe the global variability and information on decline status contained in the profiles. Two-dimensional score plots showed that participants could not be separated according to cognitive decline status, indicating that the main variability is independent from cognitive status (Figure 1A-B). The supervised PLS-DA method was used to model the relationship between the decline status and spectral profiles. No significant model PLS-DA could fit data, meaning that neither the aqueous nor the lipidomic dataset contained any predictive signature of the decline status.

[0082] Integration of the metabolomic datasets allows prediction of future cognitive decline

[0083] The two omics datasets contained complementary information and can provide a more comprehensive and detailed understanding of the metabolome, therefore we next tested whether a predictive signature of the decline status could be defined from statistically integrating both datasets. We applied the multi-omics integrative Data integration Analysis for Biomarker discovery using Latent components (DIABLO) method on Pareto-scaled data. This method seeks for common information across different data types through the selection of a subset of molecular features, while discriminating between multiple phenotypic groups (Singh et al. 2019). Bootstrap resampling method was used to assess robustness of the number of latent components and the number of variables to select per block. These optimal numbers were chosen to minimize the 4-fold cross validation-based BER. Models with 1 latent component including 1 to 20 NMR variables and 1 to 2 MS variables (data not shown) were tested to select the optimal number of variables to select in the NMR and in the MS data sets. We observed that for most of the bootstrap samples less than 20 variables in NMR block (53%) were selected and two variables in the lipidomic block (40%). Sphingomyelin d 18.1 C26.0 and triglyceride 48.3 were the most frequently selected in the models adjusted on bootstrap samples (60%, data not shown). This means that these two lipids (sphingomyelin dl8.1_C26.0 and triglyceride 48.3) are the most predictive of the decline status. Ultimately, 12 variables, including 10 from the NMR dataset and 2 from the MS dataset, together represented the best possible combination to predict decline (Figure 2A-B), and enabled discrimination of patients according to the decline status (Figure 3). These variables corresponded to seven aqueous metabolites and two lipids from 320 initial molecules (Table 2 and 3). Among the discriminating variables, four different buckets / variables were obtained for 3 hydroxybutyrate. These 9 metabolites were the most frequently selected in the models adjusted on bootstrap samples (>60%, data not shown).

[0084] The prediction model performs well in discriminating No Decline versus Decline groups

[0085] We then used receiver operating characteristic (ROC) curves to evaluate the performance of the prediction model. The results showed very good predictive ability with an AUC of 0.9167 with a p-value = 0.002028 for NMR (Figure 4A) and an AUC of 0.8021 with a p-value = 0.02526 for the lipidomic datasets (Figure 4B). We used a clustered image map to discriminate between the No Decline and Decline groups and to obtain the metabolic signature of each sample. Figure 4C shows discrimination between the two groups. These results indicate that concentrations of 3 -hydroxybutyrate, citrate, succinate, acetone, and methionine are lower in the Decline group whereas those of glucose, serine, sphingomyelin dl8.1_C26.0, and triglyceride 48.3 are higher (Table 2 and Table 3).

[0086] Conclusion

[0087] This longitudinal study uniquely provides robust predictive models of cognitive decline using untargeted metabolomics by combining both NMR and mass spectrometry and focusing on amyloid-positive individuals. Furthermore, its longitudinal nature with a relatively long follow-up and several time-points of data collection allows us to learn about the trajectories of different outcome measures of cognitive function and clearly identify a specific metabolic signature as a novel and predictive biomarker of cognitive decline.

[0088] TABLES

[0089] Table 1. Baseline characteristics of decliners (n=8) and non-decliners (n=12).

[0090] P-values determined using Fisher’s exact test for categorical variables or using Mann- Whitney U test for continuous variables. APOE, Apolipoprotein E; BMI, body mass index; CDR, Clinical Dementia Rating scale; CNT, Category Naming Test; COPD, chronic obstructive pulmonary disease; COW AT, Controlled Oral Word Association Test; FCSRT, Free and Cued Selective Reminding Test; GDS, Geriatric Depression Scale; MAPT, Multidomain Alzheimer Preventive Trial; MMSE, Mini Mental State Examination; SD, standard deviation; SPPB, Short Physical Performance Battery; SUVR = standard uptake value ratio; TMT, Trail Making Test; WAIS-R, Wechsler Adult Intelligence Scale-Revised. Table 2. Modulation of endogenous metabolites quantified in NMR spectra of plasma samples and selected by multiblock sPLS-DA. Fold change (FC) corresponds to the ratio of the mean NMR areas (Decline / No decline)

[0091] Metabolites Chemical shift FC Superpathway Subpathway

[0092] (PPm)

[0093] 3 -hydroxybutyrate 1 20 ; 2.32 ; 2.39 0.72 Lipids Ketone bodies and 2.42

[0094] Citrate 2.51 0.81 Energy TC A cycle

[0095] Succinate 2.41 0.79 Energy Succinate pathway

[0096] Acetone 2.24 0.85 Ketone bodies

[0097] Methionine 2.66 0.92 Amino acid Methionin pathway

[0098] Glucose 3.75 1.09 Carbohydrate Glycolysis, gluconeogenesis, metabolism

[0099] Serine 4.01 1.11 Amino acid Methionin pathway

[0100] Table 3. Modulation of endogenous lipids quantified by SFC-HRMS profiling in plasma samples and selected by multiblock sPLS-DA. Fold change (FC) corresponds to the ratio of the mean MS intensities (Decline / No decline)

[0101] Lipids FC Superpathway Subpathway

[0102] SM (dl8: l / C26:0) 1.535 Lipids Sphingolipid metabolism

[0103] TG (C48:3) 1.536 Lipids

[0104] Abbreviations: SM: Sphingomyelin; TG: Triglycerides.

[0105] REFERENCES:

[0106] Throughout this application, various references describe the state of the art to which this invention pertains. The disclosures of these references are hereby incorporated by reference into the present disclosure.

[0107] 1. Nichols E, Steinmetz JD, Vollset SE, Fukutaki K, Chalek J, Abd- Allah F, et al. Estimation of the global prevalence of dementia in 2019 and forecasted prevalence in 2050: an analysis for the Global Burden of Disease Study 2019. The Lancet Public Health, fevr 2022;7(2):el05-25.

[0108] 2. Mormino EC, Papp KV. Amyloid accumulation and cognitive decline in clinically normal older individuals: implications for aging and early Alzheimer’s disease. J Alzheimers Dis. 2018;64(Suppl l):S633-46. 3. van Dyck CH, Swanson CJ, Aisen P, Bateman RJ, Chen C, Gee M, et al. Lecanemab in Early Alzheimer’s Disease. N Engl J Med. 5 janv 2023;388(l):9-21.

[0109] 4. Ahadi S, Zhou W, Schiissler-Fiorenza Rose SM, Sailani MR, Contrepois K, Avina M, et al. Personal aging markers and ageotypes revealed by deep longitudinal profiling. Nat Med. janv 2020;26(l):83-90.

[0110] 5. Schiissler-Fiorenza Rose SM, Contrepois K, Moneghetti KJ, Zhou W, Mishra T, Mataraso S, et al. A longitudinal big data approach for precision health. Nat Med. mai 2019;25(5):792-804.

[0111] 6. Buergel T, Steinfeldt J, Ruyoga G, Pietzner M, Bizzarri D, Vojinovic D, et al. Metabolomic profiles predict individual multidisease outcomes. Nat Med [Internet], 22 sept 2022 [cite 20 oct 2022]; Disponible sur: https: / / www.nature.com / articles / s41591-022-01980-3

[0112] 7. Deelen J, Kettunen J, Fischer K, van der Spek A, Trompet S, Kastenmuller G, et al. A metabolic profile of all-cause mortality risk identified in an observational study of 44,168 individuals. Nat Commun. 20 aout 2019; 10(l):3346.

[0113] 8. Finkel T. The metabolic regulation ofaging. Nat Med. dec 2015;21(12): 1416-23.

[0114] 9. Ramautar R, Berger R, van der Greef J, Hankemeier T. Human metabolomics: strategies to understand biology. Current Opinion in Chemical Biology, oct 2013 ; 17(5): 841 -6.

[0115] 10. Lindon JC, Nicholson JK. The emergent role of metabolic phenotyping in dynamic patient stratification. Expert Opin Drug Metab Toxicol, juill 2014; 10(7):915-9.

[0116] 11. Dumas ME, Kinross J, Nicholson JK. Metabolic phenotyping and systems biology approaches to understanding metabolic syndrome and fatty liver disease. Gastroenterology, janv 2014;146(l):46-62.

[0117] 12. Montoliu I, Scherer M, Beguelin F, DaSilva L, Mari D, Salvioli S, et al. Serum profiling of healthy aging identifies phospho- and sphingolipid species as markers of human longevity. Aging (Albany NY), janv 2014;6(l):9-25.

[0118] 13. Proitsi P, Kim M, Whiley L, Pritchard M, Leung R, Soininen H, et al. Plasma lipidomics analysis finds long chain cholesteryl esters to be associated with Alzheimer’s disease. Transl Psychiatry. 13 janv 2015;5(l):e494.

[0119] 14. Proitsi P, Kuh D, Wong A, Maddock J, Bendayan R, Wulaningsih W, et al. Lifetime cognition and late midlife blood metabolites: findings from a British birth cohort. Transl Psychiatry. 26 sept 2018;8(l):203.

[0120] 15. Vellas B, I C, S GG, J T, T D, Jf D, et al. MAPT STUDY: A MULTIDOMAIN APPROACH FOR PREVENTING ALZHEIMER’S DISEASE: DESIGN AND BASELINE DATA. The journal of prevention of Alzheimer’s disease [Internet], juin 2014 [cite 22 mars 2023]; 1(1). Disponible sur: http: / / pubmed.ncbi.nlm.nih.gov / 26594639 /

[0121] 16. von Elm E, Altman DG, Egger M, Pocock SJ, Gotzsche PC, Vandenbroucke JP. Strengthening the reporting of observational studies in epidemiology (STROBE) statement: guidelines for reporting observational studies. BMJ. 20 oct 2007;335(7624):806-8.

[0122] 17. Giudici KV, de Souto Barreto P, Guyonnet S, Li Y, Bateman RJ, Vellas B, et al. Assessment of Plasma Amyloid-P42 / 40 and Cognitive Decline Among Community -Dwelling Older Adults. JAMA Netw Open. 1 dec 2020;3(12):e2028634.

[0123] 18. Andrieu S, Guyonnet S, Coley N, Cantet C, Bonnefoy M, Bordes S, et al. Effect of long-term omega 3 polyunsaturated fatty acid supplementation with or without multidomain intervention on cognitive function in elderly adults with memory complaints (MAPT): a randomised, placebo-controlled trial. Lancet Neurol, mai 2017;16(5):377-89.

[0124] 19. Folstein MF, Folstein SE, McHugh PR. « Mini -mental state ». A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res. nov 1975; 12(3): 189-98.

[0125] 20. Guralnik JM, Simonsick EM, Ferrucci L, Glynn RJ, Berkman LF, Blazer DG, et al. A short physical performance battery assessing lower extremity function: association with self-reported disability and prediction of mortality and nursing home admission. J Gerontol, mars 1994;49(2):M85-94.

[0126] 21. Sumner LW, Amberg A, Barrett D, Beale MH, Beger R, Daykin CA, et al. Proposed minimum reporting standards for chemical analysis Chemical Analysis Working Group (CAWG) Metabolomics Standards Initiative (MSI). Metabolomics. sept 2007;3(3):211-21.

[0127] 22. Tsugawa H, Cajka T, Kind T, Ma Y, Higgins B, Ikeda K, et al. MS-DIAL: data- independent MS / MS deconvolution for comprehensive metabolome analysis. Nat Methods, juin 2015;12(6):523-6.

[0128] 23. Le Faouder P, Soullier J, Tremblay -Franco M, Tournadre A, Martin JF, Guitton Y, et al. Untargeted Lipidomic Profiling of Dry Blood Spots Using SFC-HRMS. Metabolites. 11 mai 2021;l l(5):305.

[0129] 24. McCombie G, Browning LM, Titman CM, Song M, Shockcor J, Jebb SA, et al. omega-3 oil intake during weight loss in obese women results in remodelling of plasma triglyceride and fatty acids. Metabolomics. sept 2009;5(3):363-74. 25. Singh A, Shannon CP, Gautier B, Rohart F, Vacher M, Tebbutt SJ, et al. DIABLO: an integrative approach for identifying key molecular drivers from multi-omics assays. Bioinformatics. 1 sept 2019;35(17):3055-62.

[0130] 26. Rohart F, Gautier B, Singh A, Le Cao KA. mixOmics: An R package for ’omics feature selection and multiple data integration. PLoS Comput Biol. nov 2017; 13(1 l):el005752.

Claims

CLAIMS:

1. An in vitro method for predicting cognitive decline in a subject comprising the step of determining the level of the metabolites selected in the group consisting of 3- hydroxybutyrate, acetone, triglyceride 48:3, glucose, citrate, succinate, methionine, serine and sphingomyelin dl8: l / C26:0 in a biological sample obtained from the subject.

2. The in vitro method according to claim 1, comprising the step of: i) determining the level of the metabolites selected in the group consisting of 3- hydroxybutyrate, acetone, triglyceride 48:3, glucose, citrate, succinate, methionine, serine, sphingomyelin dl8: l / C26:0 in a biological sample obtained from the subject; ii) comparing the levels determined at step i) with their predetermined reference values and iii) concluding that the subject is at risk of suffering from cognitive decline when the levels of 3 -hydroxybutyrate, acetone, citrate, succinate and methionine determined at step i) are lower than the predetermined reference values and when the levels of glucose, serine, triglyceride 48:3 and sphingomyelin dl8: l / C26:0 determined at step i) are higher than the predetermined reference values.

3. The in vitro method according to claim 1, comprising the step of: i) determining the level of the metabolites selected in the group consisting of 3- hydroxybutyrate, acetone, triglyceride 48:3, glucose, citrate, succinate, methionine, serine, sphingomyelin dl8: l / C26:0 in a first biological sample obtained from the subject; ii) comparing the levels determined at step i) with the levels determined in a second biological sample obtained from the subject; and iii) concluding that the subject is at risk of suffering from cognitive decline when the levels of 3 -hydroxybutyrate, acetone, citrate, succinate and methionine determined at step ii) are lower than the levels determined at step i) and when the levels of glucose, serine, triglyceride 48:3 and sphingomyelin dl8: l / C26:0 determined at step ii) are higher than the levels determined at step i).

4. A method of treating cognitive decline in a subject in need thereof comprising the steps of: i) determining the level of the metabolites selected in the group consisting of 3- hydroxybutyrate, acetone, triglyceride 48:3, glucose, citrate, succinate, methionine, serine, sphingomyelin dl8: l / C26:0 in a biological sample obtained from the subject; ii) comparing the levels determined at step i) with their predetermined reference values; iii) concluding that the subject is at risk of suffering from cognitive decline when the levels of 3 -hydroxybutyrate, acetone, citrate, succinate and methionine determined at step i) are lower than the predetermined reference value and when the levels of glucose, serine, triglyceride 48:3 and sphingomyelin dl8: l / C26:0 determined at step i) are higher than the predetermined reference values; and iv) Administering said subject with a treatment against cognitive decline or practising a non-drug treatment against cognitive decline.

5. The method according to any of claims 1 to 4, wherein the subj ect suffers from attention, learning, intelligence, language, memory, perception, judgement, mental acuity, decision making, problem solving or reasoning trouble.

6. The method according to any of claims 1 to 4, wherein the subject is an amyloid positive subject.

7. The method according to any of claims 1 to 4, wherein the subject is at risk of suffering or suffers from Alzheimer’s disease.

Images