Field of the Invention

The invention relates to a prognostic method for determining the risk or likelihood of mortality in a subject suspected or diagnosed with frailty syndrome and/or response to treatment of said subject to a therapeutic treatment regimen for the treatment of same; and use of said method in a treatment regimen.

Background of the Invention

Frailty is an age-associated biological syndrome characterized by a decline in physical and mental reserves, a decrease in resistance to external stressors and an enhanced risk of disability, hospitalization and eventually death [1 , 2] The prevalence of frailty increases with age, and is typically more common in females than males, and it is estimated that in Europe frailty affects 5-10% of the population aged over 65, and between 25-50% of those older than 85 years of age.

Epidemiologic research to date has led to the identification of a number of risk factors for frailty, including: (a) chronic diseases, such as cardiovascular disease, diabetes, chronic kidney disease, depression, and cognitive impairment; (b) physiologic impairments, such as activation of inflammation and coagulation systems, anaemia, atherosclerosis, autonomic dysfunction, hormonal abnormalities, obesity, hypovitaminosis D, and environment-related factors such as life space and neighbourhood characteristics. A pro-inflammatory state, sarcopenia, anaemia, deficiencies in anabolic hormones (androgens and growth hormone) and excess exposure to catabolic hormones (cortisol), insulin resistance, compromised altered immune function, micronutrient deficiencies and oxidative stress are each individually associated with a higher likelihood of frailty. It is therefore apparent that the biological underpinnings of frailty are multifactorial, involving dysregulation across many physiological systems. To the extent that dysregulation across several physiological systems underlies the pathogenesis of frailty, specific disease states are likely concurrent manifestations of the underlying impaired physiological functions. Nevertheless, since many individuals with chronic diseases are not frail, no single disease state is necessary and sufficient to account for the pathogenesis of frailty. Therefore, diagnosing and predicting frailty and its progression, and accordingly directing treatment, can be challenging.

Further, frail elderly people are at significant risk of post-surgical complications and the need for extended care. For example, frailty more than doubles the risk of morbidity and mortality from surgery and cardiovascular conditions. Assessment of older patients before elective surgeries can accurately predict the patients' recovery trajectories. Accordingly, accurate diagnosis and prognosis of frailty syndrome is of great clinical importance.

Currently, frailty is generally assessed via an individual’s physical state and/or by the correlation with capabilities, laboratory measures and disease states that were found to associate with the syndrome in clinical practice. For example, the Fried’s criteria involves assessment of five dimensions that are hypothesized to reflect systems whose impaired regulation underlies the syndrome. These five dimensions are: unintentional weight loss, exhaustion, muscle weakness, slow gait speed, and low energy expenditure. These are typically assessed using a combination of self- reported and performance-based measures, and based on the results of these measures, individuals are divided into three categories: Robust (0 criteria), Pre-frail (1 or 2 criteria), Frail (3 or more criteria).

Another widely accepted method based on a Cumulative Deficit Model of frailty, developed as part of the Canadian Study on Health and Ageing, provides a Frailty Index. This index incorporates additional characteristics (92 parameters in the initial model, then reduced to 32) and includes abnormal laboratory values, diseases and disabilities. Flowever, in some clinical settings this model is not practical as it includes measurements of many variables.

Flowever, these current methods are somewhat imprecise or unspecific or do not provide a simple clinical means to accurately assess frailty and predict frailty-related outcomes, a requirement that is key for the determination of downstream clinical treatment. Therefore, there is an unmet need for the identification of means to assess and predict frailty-related outcomes, in particular mortality, which is of great clinical prognostic value. To this end, we herein disclose a method of stratifying frail individuals comprising detection of a biomarker to determine which of those frail individuals are at an increased risk of death, which can therefore be used to determine and direct clinical treatment pathways and downstream treatment regimen(s).

Statements of Invention

According to a first aspect of the invention there is provided a method for determining the risk or likelihood of mortality or selecting a course of treatment in a human subject suspected of having, or diagnosed with, frailty syndrome which method comprises: i. examining a biological sample from the subject to determine the amount of soluble Receptor for Advanced Glycation End-products (sRAGE), and;

ii. where the amount of sRAGE is increased compared to the amount of sRAGE in a control sample and/or the amount of sRAGE is increased compared to the amount in a reference sample and/or the amount of sRAGE is increased compared to a reference;

iii. concluding that the individual from whom the sample has been taken has a high risk or likelihood of mortality.

Reference herein to frailty syndrome refers to the age-associated biological syndrome characterized by a decline in physical and mental reserves, a decrease in resistance to external stressors and an enhanced risk of disability, hospitalization and eventually death. As will be appreciated by those skilled in the art, frailty is a multifactorial syndrome that can have complex and varied underlying etiologies but often is the consequence of a cumulative deterioration of multiple body systems and loss of their built-in reserves manifested in, for example, falls, immobility, delirium, Incontinence, susceptibility to side effects of medication. An individual having or suspected of having frailty syndrome can be diagnosed by conventional means known in the field such as, but not limited to, Fried’s five phenotypic criteria or Frailty Index as herein disclosed.

In a preferred embodiment said control sample is a sample that has been taken from a subject shown to have frailty syndrome and not at high risk of mortality and/or have an amount of sRAGE less than about 1200 pg/ml of serum, ideally, using any one or more of the above conventional techniques for identifying same. For example, this could include individuals with 3 or more of Fried’s criteria as discussed herein. Preferably said control sample is a fluid sample, such as blood including serum and/or plasma. Ideally, the sample is that of serum.

In yet a further preferred embodiment said reference sample is an amount of sRAGE in a sample from a healthy individual, which may or may not be age/gender/weight- matched with said subject, and said increase in sRAGE is a certain increase, such as a fold increase, over said reference. In an alternative embodiment said reference is a threshold value above which amounts of sRAGE are indicative of increased likelihood of mortality in an individual suspected of having, or diagnosed with, frailty.

As is known in the art, Receptor for Advanced Glycation End-products (RAGE) is a multiligand pattern recognition cell surface receptor that belongs to the immunoglobulin superfamily. RAGE binds a variety of damage-associated molecules, including glycated proteins, lipids and DNA, as well as pathogen-associated molecules from bacteria, viruses and parasites, and upon activation, it elicits pro-inflammatory processes [3] In addition to cell surface bound RAGE, there are two major soluble forms of this receptor circulating in blood, both of which lack its membrane and cytoplasmic domains, and are collectively known as soluble RAGE or sRAGE [4] One form of sRAGE, called cRAGE, is generated by the proteolytic cleavage and ectodomain shedding of cell surface RAGE. The second form, called esRAGE or RAGEvl , results from alternative splicing of RAGE mRNA and accounts for about 20% of circulating sRAGE. Although the precise function that sRAGE plays in human biology remains unresolved, there is evidence to suggest that the total amount of this molecule in the circulation reflects RAGE activation, thus potentially making it a useful biomarker of underlying inflammatory pathologies. Amounts of sRAGE have been measured in humans to search for associations with disease states or their risk factors [4, 5] With some exceptions, studies have generally reported that blood amounts of sRAGE were elevated in diabetics and in people with kidney disease. In contrast, in other chronic or inflammatory conditions sRAGE amounts were reported to be lower than in healthy subjects.

Therefore, as will be appreciated, detecting sRAGE can involve the detection of cRAGE, and/or esRAGE, and variants thereof. Reference herein to a biological sample refers to a sample taken from a subject who has or is suspected of suffering from frailty syndrome. In a further preferred method of the invention said biological sample is a fluid. Ideally, the biological sample is a blood sample, including serum and/or plasma.

As will be appreciated by those skilled in the art, expression of sRAGE can be determined by numerous means including, but not limited to, detection of sRAGE protein, or fragments thereof. Alternatively, expression of RAGE mRNA including variants of RAGE such as esRAGE, or cDNA can be undertaken by conventional means to indicate the presence of sRAGE protein.

In a preferred method of the invention the method comprises the detection of sRAGE protein and comprises the following steps:

a) providing a biological sample from a subject to be analyzed;

b) forming a preparation comprising said biological sample and one or more antibodies, that specifically bind sRAGE polypeptide(s) or peptide fragments thereof, to form respectively antibody/sRAGE complexes or antibody/peptide fragment complexes;

c) detecting the complexes which are indicative of the amount of sRAGE; d) comparing the amount of sRAGE in the subject sample to a control sample or a reference sample or a reference; and

e) where the amount of sRAGE is increased in the subject sample compared to the control sample or the reference sample or the reference concluding that the individual from whom the sample has been taken has a high risk or likelihood of mortality.

In a preferred embodiment, said sRAGE protein(s) comprises the amino acid sequence set forth in SEQ ID NO: 1 and/or SEQ ID NO: 2 or fragments thereof, for esRAGE and cRAGE, respectively:

SEQ ID NO: 1

M AAGTAVG AWVLVLS LWG AWGAQ N ITAR I G E P LVL KC KGAP KKPPQRLEWKLNT GRTEAWKVLSPQGGGPWDSVARVLPNGSLFLPAVGIQDEGIFRCQAMNRNGKET KSNYRVRVYQIPGKPEIVDSASELTAGVPNKVGTCVSEGSYPAGTLSWHLDGKPLV PNEKGVSVKEQTRRHPETGLFTLQSELMVTPARGGDPRPTFSCSFSPGLPRHRAL

RTAPIQPRVWEPVPLEEVQLWEPEGGAVAPGGTVTLTCEVPAQPSPQIHWMKDG

VPLPLPPSPVLILPEIGPQDQGTYSCVATHSSHGPQESRAVSISIIEPGEEGPTAGE G

FDKVREAEDSPQHM

SEQ ID NO: 2.

M AAGTAVG AWVLVLS LWG AWGAQ N ITAR I G E P LVL KC KGAP KKPPQRLEWKLNT

GRTEAWKVLSPQGGGPWDSVARVLPNGSLFLPAVGIQDEGIFRCQAMNRNGKET

KSNYRVRVYQIPGKPEIVDSASELTAGVPNKVGTCVSEGSYPAGTLSWHLDGKPLV

PNEKGVSVKEQTRRHPETGLFTLQSELMVTPARGGDPRPTFSCSFSPGLPRHRAL

RTAPIQPRVWEPVPLEEVQLWEPEGGAVAPGGTVTLTCEVPAQPSPQIHWMKDG

VPLPLPPSPVLILPEIGPQDQGTYSCVATHSSHGPQESRAVSISIIEPGEEGPTAG

In a preferred method of the invention said antibody detection involves use of an immunoassay, for example an ELISA.

In this embodiment of the invention the method involves assaying for the sRAGE protein or at least one peptide fragment thereof and typically, but not exclusively, it involves the use of agents that bind to the relevant proteins and so identify the sRAGE protein or peptide fragment. Common agents are for example antibodies of which there are examples in the art, e.g. [17] or [18] or Quantikine Human RAGE Immunoassay, and most ideally, monoclonal antibodies which, advantageously, have been labelled with a suitable tag whereby the existence of the bound antibody can be determined. Assay techniques for identifying proteins are well known to those skilled in the art and, indeed, used every day by workers in the field of clinical diagnostics. Such assay techniques may be applied by the skilled worker to work the invention.

In an alternative preferred method of the invention said method detects RAGE mRNA or splice variants thereof or cDNA.

In this preferred method of the invention said method comprises:

a) providing a biological sample taken from a subject to be analyzed;

b) forming a preparation comprising said biological sample and one or more nucleic acid probes complementary to all or part of a nucleic acid encoding RAGE;

c) providing conditions to detect, or optionally amplify, a nucleic acid molecule encoding RAGE in said biological sample; d) quantifying the amount of RAGE nucleic acid expression in said biological sample;

e) comparing the amount of RAGE nucleic acid expression in the subject sample compared to a control sample or a reference sample or a reference; and f) where the amount of RAGE nucleic acid expression in the subject sample is increased compared to the amount in the control sample or the reference sample or the reference concluding that the individual from whom the sample has been taken has a high risk or likelihood of mortality.

In a preferred embodiment, said RAGE nucleic acid(s) comprises the nucleic acid sequence set forth in SEQ ID NO: 3 or SEQ ID NO: 4, for esRAGE and RAGE, respectively:

SEQ ID NO: 3

ctgaaagatgggggctggagagagggtgcaggccccacctagggcggaggccacagc agggagaggggcag acagagccaggaccctggaaggaagcaggatggcagccggaacagcagttggagcctggg tgctggtcctcagt ctgtggggggcagtagtaggtgctcaaaacatcacagcccggattggcgagccactggtg ctgaagtgtaaggggg cccccaagaaaccaccccagcggctggaatggaaactgaacacaggccggacagaagctt ggaaggtcctgtct ccccagggaggaggcccctgggacagtgtggctcgtgtccttcccaacggctccctcttc cttccggctgtcgggatcc aggatgaggggattttccggtgccaggcaatgaacaggaatggaaaggagaccaagtcca actaccgagtccgtg tctaccagattcctgggaagccagaaattgtagattctgcctctgaactcacggctggtg ttcccaataaggtggggac atgtgtgtcagagggaagctaccctgcagggactcttagctggcacttggatgggaagcc cctggtgcctaatgaga agggagtatctgtgaaggaacagaccaggagacaccctgagacagggctcttcacactgc agtcggagctaatgg tgaccccagcccggggaggagatccccgtcccaccttctcctgtagcttcagcccaggcc ttccccgacaccgggcc ttgcgcacagcccccatccagccccgtgtctgggagcctgtgcctctggaggaggtccaa ttggtggtggagccaga aggtggagcagtagctcctggtggaaccgtaaccctgacctgtgaagtccctgcccagcc ctctcctcaaatccactg gatgaaggatggtgtgcccttgccccttccccccagccctgtgctgatcctccctgagat agggcctcaggaccaggg aacctacagctgtgtggccacccattccagccacgggccccaggaaagccgtgctgtcag catcagcatcatcgaa ccaggcgaggaggggccaactgcaggtgaggggtttgataaagtcagggaagcagaagat agcccccaacaca tgtgactggggggatggtcaacaagaaaggaatggaaggccccagaaaaccaggaggaag aggaggagcgtg cagaactgaatcagtcggaggaacctgaggcaggcgagagtagtactggagggccttgag gggcccacagaca gatcccatccatcagctcccttttctttttcccttgaactgttctggcctcagaccaact ctctcctgtataatctctctcctgtat aaccccaccttgccaagctttcttctacaaccagagccccccacaatgatgattaaacac ctgacacatcttgcaaaa aaaaaaaaaaaaaa

SEQ ID NO: 4:

ctgaaagatgggggctggagagagggtgcaggccccacctagggcggaggccacagc agggagaggggcag acagagccaggaccctggaaggaagcaggatggcagccggaacagcagttggagcctggg tgctggtcctcagt ctgtggggggcagtagtaggtgctcaaaacatcacagcccggattggcgagccactggtg ctgaagtgtaaggggg cccccaagaaaccaccccagcggctggaatggaaactgaacacaggccggacagaagctt ggaaggtcctgtct ccccagggaggaggcccctgggacagtgtggctcgtgtccttcccaacggctccctcttc cttccggctgtcgggatcc aggatgaggggattttccggtgccaggcaatgaacaggaatggaaaggagaccaagtcca actaccgagtccgtg tctaccagattcctgggaagccagaaattgtagattctgcctctgaactcacggctggtg ttcccaataaggtggggac atgtgtgtcagagggaagctaccctgcagggactcttagctggcacttggatgggaagcc cctggtgcctaatgaga agggagtatctgtgaaggaacagaccaggagacaccctgagacagggctcttcacactgc agtcggagctaatgg tgaccccagcccggggaggagatccccgtcccaccttctcctgtagcttcagcccaggcc ttccccgacaccgggcc ttgcgcacagcccccatccagccccgtgtctgggagcctgtgcctctggaggaggtccaa ttggtggtggagccaga aggtggagcagtagctcctggtggaaccgtaaccctgacctgtgaagtccctgcccagcc ctctcctcaaatccactg gatgaaggatggtgtgcccttgccccttccccccagccctgtgctgatcctccctgagat agggcctcaggaccaggg aacctacagctgtgtggccacccattccagccacgggccccaggaaagccgtgctgtcag catcagcatcatcgaa ccaggcgaggaggggccaactgcaggctctgtgggaggatcagggctgggaactctagcc ctggccctggggatc ctgggaggcctggggacagccgccctgctcattggggtcatcttgtggcaaaggcggcaa cgccgaggagaggag aggaaggccccagaaaaccaggaggaagaggaggagcgtgcagaactgaatcagtcggag gaacctgaggc aggcgagagtagtactggagggccttgaggggcccacagacagatcccatccatcagctc ccttttctttttcccttga actgttctggcctcagaccaactctctcctgtataatctctctcctgtataaccccacct tgccaagctttcttctacaacca gagccccccacaatgatgattaaacacctgacacatcttgcaaaaaaaaaaaaaaaaaa

As will be appreciated by those skilled in the art, detection of RAGE nucleic acid in said subject sample can be achieved by using numerous nucleic acid detection techniques such as, but not limited to, nucleic acid binding assays, including polymerase chain reaction [PCR] based methods, or the like. The detection methods described above may be qualitative and/or quantitative. However, other conventional techniques, as will be appreciated by those skilled in the art, can be used in accordance with the invention.

In a preferred method of the invention said method is a PCR based method.

In yet a preferred method of the invention said PCR based method is Real Time [RT] PCR including real-time quantitative PCR.

Ideally, the sample that is examined is assayed for the presence of RNA, preferably total RNA and, more preferably still, the amount of mRNA encoding RAGE. It will be apparent to those skilled in the art that techniques available for measuring RNA content are well known and, indeed, routinely practised by those in the clinical diagnostics field. Such techniques may include reverse transcription of RNA to produce cDNA and an optional amplification step followed by the detection of the cDNA or a product thereof.

In a preferred embodiment, as will be appreciated by those skilled in the art, detection is most ideally practised using isoform specific nucleic acid probes that can specifically and accurately detect, and preferably allow quantification of the amount of RAGE, including esRAGE, or variants thereof.

As used herein, the term "probe(s)" describes an oligonucleotide that hybridizes under physiological or reaction conditions to RAGE target cDNA or RNA. Those skilled in the art will recognize that the exact length of the oligonucleotide and its degree of complementarity with its target will depend upon the specific target selected, including the sequence of the target and the particular bases which comprise that sequence.

It is preferred that the probe be constructed and arranged so as to bind selectively with the target under physiological/reaction conditions, i.e. , to hybridize substantially more to the target sequence than to any other sequence in the sample under physiological/reaction conditions.

In order to be sufficiently selective and potent the probes comprise at least 7 and more preferably, at least 15 consecutive bases which are complementary to the target. Most preferably, the probes comprise a complementary sequence of 20-30 bases.

In a preferred method of the invention said method involves the detection of RAGE nucleic acid expression or sRAGE protein in the said sample, which detection is converted into an amount of RAGE nucleic acid expression or sRAGE protein in said sample and then the amount is compared with the amount of RAGE nucleic acid expression or sRAGE protein in a control sample or reference sample or reference and where the amount of sRAGE protein or RAGE nucleic acid in said sample is greater than that in the control sample or reference sample or reference concluding the existence of mortality linked to frailty syndrome.

Increased RAGE nucleic acid expression in a sample correlates with an increased sRAGE protein; either of which can be compared to the control sample or reference sample or reference. In all cases the normal, increased or decreased expression is statistically relevant at the 5% level or less.

We have unexpectedly found that increased sRAGE at the nucleic acid or protein level in subjects with, or suspected of having, frailty correlates with poor survival and an increased likelihood of mortality. In particular, in frail individuals with low or medium levels of sRAGE it was found that the incidence of mortality in a 6 year monitoring or follow-up period was reduced compared to frail individuals with high levels of sRAGE. Therefore, increased expression of the sRAGE is indicative of a poorer overall survival.

In a preferred embodiment, the threshold value above which amounts of sRAGE are indicative of a high risk of mortality in an individual suspected of having, or diagnosed with, frailty is an amount of at least 1 195 pg/ml of serum or about at least 1200 pg/ml of serum. More preferably the threshold value is an amount of at least 1294pg/ml of serum or about at least 1300 pg/ml of serum, more preferably still an amount of at least 1626 pg/ml of serum or about at least 1600 pg/ml of serum, and most ideally at an amount of at least 1887 pg/ml of serum or an amount of about at least 1900 pg/ml of serum.

In a preferred embodiment, the threshold value above which amounts of sRAGE are indicative of a high risk of mortality in an individual suspected of having, or diagnosed with, frailty is an amount selected from at least: 1200pg/ml, 1250pg/ml, 1300pg/ml, 1350pg/ml, 1400pg/ml, 1450pg/ml, 1500pg/ml, 1550pg/ml, 1600pg/ml, 1650pg/ml, 1700pg/ml, 1750pg/ml, 1800pg/ml, 1850pg/ml, 1900pg/ml and every 1 pg/ml therebetween.

Thus for example, an individual’s chances of dying at any given time within a four year follow up period, increases on average 4.5-fold if the amount of sRAGE is about at least 1300 pg/ml and up to at least 6-fold if the amount of sRAGE is about at least 1900 pg/ml, in each case compared to an individual’s chances of dying in a suitable reference group, even after adjustment for potential confounders. Likewise, an individual’s chances of dying at any given time within a six year follow up period, increases on average 2.5-fold if the amount of sRAGE is about at least 1300 pg/ml and up to at least 2.8-fold if the amount of sRAGE is about at least 1900 pg/ml, in each case compared to an individual’s chances of dying in a suitable reference group, even after adjustment for potential confounders. This implies sRAGE is a remarkable predictive marker for determining mortality in individuals with frailty syndrome. Thus, stratifying frail individuals into groups according to cut-offs values of sRAGE, for example, below or above the median, by tertiles or by quartiles of sRAGE, means we can determine which frail individuals will survive for longer periods of time based on the amount of sRAGE they have; this type of stratification does not provide any additional prognostic information for non-frail individuals.

In yet a further preferred embodiment, said method for determining the risk or likelihood of mortality in a human subject suspected of having, or diagnosed with, frailty syndrome further comprises selecting a course of treatment for said subject based upon the sRAGE test described herein, thus the method comprises performing any one of the afore methods and then, depending upon the outcome of the method, determining a suitable or selected course of treatment. For example, treatment may include lifestyle intervention such as, but not limited to, diet modifications aimed at reducing the amount of Advanced glycation-end products in blood (e.g. steaming food instead of baking or roasting) or promoting a more active lifestyle including physical exercise which is known to fight inflammation. Additionally, or alternatively, stratification may permit utilization of certain therapeutic interventions including but not limited to RAGE inhibitors, metformin or senolytics, the latter being a group of anti- ageing substances that target and eliminate senescent cells. Further, as alluded to, such a prognostic also permits stratification of frail individuals that should, or should not, undergo certain surgical or clinical intervention(s), such as but not limited to, frail individuals with aortic stenosis and whether they can undergo conventional surgery versus the less risky (yet more costly) transcathether aortic valve implantation, or alternatively not undergo surgery at all. Further, frailty is strongly associated with mortality risk in individuals suffering from cancer who undergo palliative chemotherapy, and therefore the methodology may be used to stratify frail individuals suffering from cancer that should or should not receive palliative chemotherapy, for example frail patients with colorectal cancer.

According to a further aspect of the invention, there is provided a method to determine if a subject suspected or diagnosed with frailty syndrome will or will not respond to a selected course of treatment (for a condition such as by way of example frailty syndrome) comprising determining the amount of nucleic acid expression encoding RAGE or the amount of sRAGE protein in a plurality of biological samples taken periodically from a human subject and wherein changes in the amount of RAGE nucleic acid or sRAGE protein determines a treatment regimen for said subject.

In a preferred method of the invention the method comprises the detection of sRAGE protein and comprises the following steps:

a) providing a biological sample from a subject to be analyzed;

b) forming a preparation comprising said biological sample and one or moreantibodies that specifically bind sRAGE protein or a peptide fragment thereof, to form antibody/protein complexes or antibody/peptide complexes;

c) detecting the complexes which are indicative of the amount of sRAGE;

d) comparing the amount of sRAGE in the subject sample to a control sample or a reference sample or a reference; and

e) where the amount of sRAGE is increased in the subject sample compared to the control sample or the reference sample or the reference concluding that the individual from whom the sample has been taken has a high risk or likelihood of mortality;

f) periodically repeating step a) to e); and

g) where changes in the amount of sRAGE are determined under parts d) or e) are identified adjusting the subject’s treatment.

In an alternative preferred method of the invention said method comprises the detection of RAGE nucleic acid expression and comprises the following steps:

a) providing a biological sample taken from a subject to be analyzed;

b) forming a preparation comprising said biological sample and one or more nucleic acid probes complementary to all or part of the nucleic sequence encoding RAGE;

c) providing conditions to detect, or optionally amplify, a nucleic acid molecule expressed and encoding RAGE in said biological sample;

d) quantifying the amount of expression of RAGE in said biological sample;

e) comparing the amount of RAGE in the subject sample to a control sample or a reference sample or a reference and where the amount of RAGE in the subject sample is increased compared to the amount in the control sample or the reference sample or the reference concluding that the individual from whom the sample has been taken has a high risk or likelihood of mortality; and

f) periodically repeating step a) to e) and where changes in the amount of RAGE determined under parts d) or e) are identified adjusting the subject’s treatment.

We have discovered that amounts of sRAGE protein in a sample correlate with known frailty syndrome staging techniques such that a simple in vitro assay can be used to reliably inform a clinician about, not only the existence of frailty syndrome, but also its likely progression. Thus, in a further preferred method of the invention, one assess whether sRAGE is progressively increasing over time or is sustained at as an increased amount over time as this too is indicative of subjects with a high risk mortality.

As will be appreciated by those skilled in the art, the observation that sub-groups of subjects with frailty syndrome with increased sRAGE expression at either the nucleic acid level or protein level, have a poor prognosis and disease outcome can be used to stratify patients with a view to determining those likely to respond to clinical or therapeutic intervention such that informed treatment decisions can be made. More preferably, as it has been determined that increased sRAGE expression is indicative of poorer prognosis, stratification of those patients groups most likely to respond to a selected course of clinical treatment can be made.

According to a yet further aspect of the invention there is provided a method of treating a subject suspected of having, or diagnosed with, frailty syndrome comprising:

a) obtaining a biological sample from said subject to be analysed;

b) quantifying the amount of nucleic acid expression encoding RAGE or the amount of sRAGE protein in said biological sample; and

c) where the amount of either of said RAGE or sRAGE is increased compared to the amount of the corresponding amount of said RAGE or sRAGE in a control sample and/or the amount of either of said RAGE or sRAGE is increased compared to the amount in a reference sample and/or the amount of either of said RAGE or sRAGE is increased compared to a reference undertaking a suitable or selected course of treatment. Throughout the description and claims of this specification, the words“comprise” and “contain” and variations of the words, for example“comprising” and“comprises”, mean “including but not limited to” and do not exclude other moieties, additives, components, integers or steps. Throughout the description and claims of this specification, the singular encompasses the plural unless the context otherwise requires. In particular, where the indefinite article is used, the specification is to be understood as contemplating plurality as well as singularity, unless the context requires otherwise.

All references, including any patent or patent application, cited in this specification are hereby incorporated by reference. No admission is made that any reference constitutes prior art. Further, no admission is made that any of the prior art constitutes part of the common general knowledge in the art.

Preferred features of each aspect of the invention may be as described in connection with any of the other aspects.

Other features of the present invention will become apparent from the following examples. Generally speaking, the invention extends to any novel one, or any novel combination, of the features disclosed in this specification (including the accompanying claims and drawings). Thus, features, integers, characteristics, compounds or chemical moieties described in conjunction with a particular aspect, embodiment or example of the invention are to be understood to be applicable to any other aspect, embodiment or example described herein, unless incompatible therewith.

Moreover, unless stated otherwise, any feature disclosed herein may be replaced by an alternative feature serving the same or a similar purpose.

The Invention will now be described by way of example only with reference to the Examples below and to the following Figures and Tables wherein:

Figure 1. Flow chart for study sample selection. Abbreviations: FC, Fried’s criteria;

Figure 2. Comparison of baseline sRAGE levels between 4-year survivors and non- survivors according to their frailty status; Figure 3. Kaplan Meier survival curves of frail participants by sRAGE levels below and above the median (A), tertiles (B) and quartiles (C);

Figure 4. Kaplan Meier survival curves of non-frail participants by sRAGE levels below and above the median (A), tertiles (B) and quartiles (C);

Figure 5. Integrated Areas Under the Curve (AUC) for different time horizons.

Values were calculated for hazard models which included or excluded sRAGE as indicated;

Table 1. Baseline sociodemographic, behavioural and health characteristics of the study participants. Abbreviations: BMI, body mass index; ADL, activities of daily living; sRAGE, soluble receptor for advanced glycation-end products;

Table 2. Cumulative deaths after each follow-up year;

Table 3. Univariate analysis for potential contributing factors of mortality (4 years follow-up). The unadjusted hazard ratios (HR) for death (with their respective 95% confidence intervals and P values) for each of the variables included in the study are given for the entire analytical sample (all) and for the non-frail and frail subsets within the sample;

Table 4. Multivariate Cox proportional hazard models of the relationship between Ln sRAGE and mortality (4 years follow-up). The unadjusted and adjusted hazard ratios (HR) for death per unit increment of the natural log-transformed sRAGE (with their respective 95% confidence intervals and P values) are given for the entire analytical sample (all) and for the non-frail and frail subsets within the sample. Model 1 = Multivariate model adjusted for cohort; Model 2 = Model 1 additionally adjusted for age and gender; Model 3 = Model 2 additionally adjusted for smoking history, total cholesterol, creatinine, ADL, diabetes, cardiovascular disease and stroke; Model 4 = Model 3 additionally adjusted for cancer; Table 5. Multivariate Cox proportional hazard model of the relationship between LnsRAGE and mortality (4, 5 and 6-years follow-up). The multivariate model corresponds to model 4 in Table 4;

Table 6. Multivariate Cox proportional hazard models of the relationship between sRAGE and mortality in frail participants according to sRAGE levels below or above the median, tertiles and quartiles (4 years follow-up). Model 1 = Multivariate model adjusted for cohort; Model 2 = Model 1 additionally adjusted for age and gender; Model

3 = Model 2 additionally adjusted for smoking history, total cholesterol, creatinine, ADL, diabetes, cardiovascular disease and stroke; Model 4 = Model 3 additionally adjusted for cancer;

Table 7. Multivariate Cox proportional hazard model of the relationship between sRAGE and mortality in frail participants according to sRAGE levels below and above the median, tertiles and quartiles (4, 5 and 6 years follow-up). The multivariate model corresponds to model 4 in Table 6;

Table 8. Multivariate Cox proportional hazard models of the relationship between Ln sRAGE and mortality (6 years follow-up). The unadjusted and adjusted hazard ratios (HR) for death per unit increment of the natural log-transformed sRAGE (with their respective 95% confidence intervals and P values) are given for the entire analytical sample (All) and for the non-frail and frail subsets within the sample. Models correspond to those in Table 4, except that Model 3 and Model 4 were adjusted for eGFR instead of creatinine;

Table 9. Multivariate Cox proportional hazard models of the relationship between sRAGE and mortality in frail participants according to sRAGE quartiles (6 years follow- up). Models correspond to those in Table 8; and

Table 10. Comparison of the relationship between sRAGE and mortality in frail and non-frail individuals based on analytical sample quartiles (6 years follow-up). The multivariate model was adjusted for cohort, age, gender, smoking history, total cholesterol, eGFR, ADL, diabetes, cardiovascular disease, stroke and cancer (model

4 in Table 8). MATERIALS AND METHODS

Participants

Participants in this study were men and women aged 65 and older from two well characterised population-based European cohorts, namely, Toledo Study of Healthy Ageing (TSHA) [6] and Approche Multidisciplinaire Integree (AMI) [7], who were enrolled in 2013 as part of the exploratory phase of the FRAILOMIC initiative, a European project investigating biomarkers of frailty [8]

The TSHA is a prospective cohort study initiated in 2006 aimed at studying the determinants and consequences of frailty in community-dwelling elderly individuals living in the city of Toledo and neighbouring towns, Spain. AMI is a population-based prospective cohort started in 2007 to study health and ageing in elderly farmers living in rural South West France. The complete methodologies for recruitment and investigations of participants from the two cohorts have been reported elsewhere [6, 7, 9] The TSHA study protocol was approved by the Clinical Research Ethics Committee of the Complejo Hospitalario de Toledo (Spain) and the AMI study was approved by the Ethics Committee of the CHU (University Hospital) of Bordeaux (France). The research followed the principles embodied in the Declaration of Helsinki.

In both cohorts, once participants gave written informed consent, biological samples and a wide range of sociodemographic, behavioural and health-related baseline data were collected between 2006 and 2009 by trained psychologists and/or nurses during home visits. Individuals were considered for inclusion in FRAILOMIC if a stored sample of plasma, serum and/or urine was available for biomarker evaluation and if the frailty status could be determined from case report forms using the frailty definition proposed by Fried et al. [1 ] (see below). A total of 1398 participants from TSHA and 695 from AMI fulfilled the aforementioned selection criteria. To ensure adequate number of participants for statistical analysis, enrolment of the FRAILOMIC participants was carried out to achieve an approximate ratio 1 case to 3 controls and a similar cardiovascular risk profile in both groups. This resulted in 474 subjects from TSHA (109 frail and 365 non-frail) and 320 subjects from AMI (80 frail and 240 non- frail) in being included in the FRAILOMIC database. Of these, 48 participants in TSHA and 2 in AMI who did not have enough stored serum to measure sRAGE were excluded from the present study. Similarly, 3 participants from TSHA and 30 from AMI whose case report forms had missing data on 1 or 2 of the 5 Fried’s frailty parameters were also excluded. This left a combined TSHA-AMI sample pool of 71 1 participants to be considered for analysis in the current study.

Measurement of frailty

Frailty was evaluated using Fried’s frailty phenotype [1 ], which includes five criteria, namely slow walking speed, weakness, weight loss, self-reported exhaustion, and low physical activity. 1 ) Slow walking speed was defined as the worst quintile in a 3-m walking speed test adjusted for gender and height. 2) Weakness was defined as the lowest quintile of grip strength measured with a Jamar hand dynamometer, after adjustment for gender and BMI (in kg/m ² ) for TSHA [10] or as having difficulty rising from a chair without using armrests in AMI [1 1 ] 3) Weight loss, was defined as the unintentional loss of at least 4.5 kg in the preceding year for TSHA or >3 kg in the previous 3 months for AMI. 4) Self-reported exhaustion was evaluated in both cohorts based on a positive response to any of the following questions from the Center for Epidemiologic Studies Depression Scale [12]:“I felt that anything I did was a big effort” or“I felt that I could not keep on doing things”“at least 3-4 days a week”. 5) Low physical activity was defined as the lowest quintile for each gender of the Physical Activity Scale for the Elderly [13] in the case of TSHA, and as <1 hour of exercise/week or <3.5 hours of leisure activities/week for AMI. For the purpose of this study individuals meeting three or more criteria were classed as frail and those that met none, one or two criteria were classed as non-frail.

Measurement of sRAGE

Serum levels of sRAGE were determined from fasting blood samples stored at -80°C using a commercially available sandwich ELISA which detects both, cRAGE and esRAGE (Quantikine Human RAGE Immunoassay, R&D Systems, Abingdon, UK). All measurements were carried out in the same laboratory with the nature and origin of the samples blinded to the operator. The intra- and inter-assay coefficients of variation were 1 .7% and 3.7%, respectively. Other variables

Sociodemographic and behavioural information recorded at baseline included age, gender, level of education and smoking history. Participants were also asked to report whether they had previously suffered from any of the following physician-diagnosed diseases: hypertension, diabetes mellitus, cardiovascular disease (ischemic heart disease or heart failure), stroke or cancer. Body Mass Index (BMI) was calculated as weight in kilograms divided by height in meters squared. Limitations in basic activities of daily living (ADLs) were measured with the use of the Katz ADL Scale [14] Obesity was defined according to World Health Organization guidelines. Cholesterol and creatinine was measured from fasting blood samples by routine enzymatic methods. The estimated glomerular filtration rate (eGFR) was calculated using the CKD-EPI creatinine equation [15]

Mortality

Dates of death were obtained from the Spanish National Death Index (Ministry of Health and Social Services) for TSHA participants and from the death registries of regional Councils for AMI participants. Where necessary in both cohorts deaths were confirmed by follow-up telephone interviews with relatives of the deceased. The date of death was recorded up until 6 years from baseline, at which point the study was right-censored.

Statistical analysis

For comparison of sociodemographic characteristics and health indicators continuous variables are reported as means ± SD for normally distributed data or as medians with IQR for skewed data. Categorical variables are presented as percentages. Differences in characteristics between frail and non-frail groups were compared by c2 test for categorical variables, and by Student’s t test or Mann-Whitney test for continuous variables, as appropriate.

Cox proportional hazards models examining the associations between sRAGE and mortality were performed with SSPS v.23 for Windows (SSPS, Chicago, USA). Given the skewed distribution of sRAGE, for its evaluation as a continuous variable it was natural log-transformed before further analysis. Analyses were carried out for the full analytical sample and then repeated for the frail and non-frail groups. Multivariate models were adjusted for those variables identified by univariate analysis as conferring a statistically significant risk of death (P<0.05) in the full analytical sample. In addition, models were adjusted for cohort origin and history of cancer. For the main analyses the associations were estimated based on a follow-up period of four years. Analyses were also performed using follow-up periods of five and six years, and where indicated creatinine was replaced for eGFR as a covariate.

Cumulative survival time was calculated by the Kaplan-Meier method and analyzed by the log-rank test using Graph Pad Prism (v.5.01 ).

The estimated integrated area under curve (IAUC) was calculated over a time frame of 2 to 6 years of follow-up using the Statistical package R v.2.15.2 for Windows (Vienna, Austria). The IAUC was plotted against time using Graph Pad Prism (v. 5.01 ).

RESULTS

Comparison between frail and non-frail participants

Figure 1 shows a flow chart depicting the selection of participants from the TSFIA and AMI cohorts enrolled in FRAILOMIC and their progression through to the current study. We measured sRAGE in serum samples of 71 1 FRAILOMIC participants who had full frailty data. Of these, we excluded 16 participants from further analysis for missing covariates at baseline. We also excluded 4 additional participants for giving extremely high values of sRAGE (over 4 standard deviations above the mean), thus leaving a total of 691 participants in the analytical sample.

Table 1 shows the baseline demographic and health characteristics of the participants included in the analytical sample and their comparison according to frailty status. One hundred and forty-one participants (20.4%) were considered to be frail at baseline, 50 (35.5%) of which were from AMI and 91 (64.5%) were from TSFIA. Frail participants were older and had a lower level of education than their non-frail counterparts, with a larger proportion of them being females. Frail participants had a higher BMI, higher rates of obesity and a higher dependence for basic ADLs. Frail participants showed also a higher prevalence of comorbidities, including diabetes mellitus, cardiovascular diseases and stroke, but not of cancer or hypertension. Table 1 also shows that frail participants had higher concentrations of sRAGE than the non-frail, although the difference did not reach statistical significance. Table 2 shows that during follow-up periods of up to six years a larger proportion of participants died within the frail group than within the non-frail group.

To examine the relationship between sRAGE and mortality, analyses were initially performed based on four years follow-up data. As shown in Figure 2, in frail individuals who had died within this period, median sRAGE levels at baseline were significantly higher than in those that survived (1959 [1 195-2674] pg/mL vs 1213 [871 -1683] pg/mL, P<0.001 ). In contrast, no difference in sRAGE levels was seen between non-survivors and survivors of the non-frail group (1245 [998-1705] pg/mL vs 1 186 [921 -1550] pg/mL, P= 0.337). Comparable results were obtained when this analysis was repeated based on 6 years follow-up data. In this case in the frail group median sRAGE levels at baseline were 1563 [1015-2248] pg/mL in individuals who had died compared to 1 184 [870-1657] pg/mL in those who had survived, this difference being significant (P=0.006), whereas within the same monitoring period no difference in sRAGE levels was seen between non-survivors and survivors of the non-frail group (1262 [1056- 1554] pg/mL vs 1 186 [919-1551 ] pg/mL, P=0.19). sRAGE as a prognostic marker of mortality in frail individuals - Results from univariate Cox proportional hazards regression models

As shown in Table 3, in univariate Cox proportional hazards regression models sRAGE (entered as a natural log-transformed continuous variable), was a significant predictor of mortality when all the participants in the sample were considered together (HR=2.76, 95%CI 1 .61 -4.73, P <0.001 ). Strikingly, when the analytical sample was divided into non-frail and frail participants, sRAGE remained a significant predictor of mortality in frail participants (HR=4.18, 95%CI 1 .91 -9.16, P <0.001 ), but not in those that were classified as non-frail (HR=1 .55, 95%CI 0.75-3.23, P <0.238). Several other characteristics were also associated with mortality in the full analytical sample (Table 3). These included age, male gender, smoking history, total cholesterol, creatinine, inability to perform basic ADLs, frailty and a diagnosis of diabetes mellitus, cardiovascular disease or stroke. Most of these associations were also maintained in the frail group. However, unlike in the case of sRAGE, and except for diabetes mellitus, all these variables also remained associated with mortality in the non-frail group. sRAGE as a prognostic marker of mortality in frail individuals - Results from multivariate Cox proportional hazards regression models

To examine the influence of potential confounding factors on the association between sRAGE and mortality, those variables which showed a significant association with mortality for all the participants in the univariate model, as well as two other relevant variables, namely the cohort origin and cancer, were entered into a multivariate Cox proportional hazard regression analysis. As shown in Table 4, four stepwise models with different sets of variables were fitted. Consistent with the results of the unadjusted models, the association between sRAGE and mortality significantly persisted across all the adjusted models for all the participants in the sample and also for the frail group, but was absent in the non-frail group. Notably, in all the models the estimated HR of death conferred by sRAGE were always larger in the frail group than in all the participants, suggesting that the association between sRAGE and death observed in the full analytical sample, is due to the contribution of the frail participants in the sample. Concerning the frail group, Model 1 demonstrated that the cohort origin was not a confounding factor. Model 2 showed that further adjustment for age and gender attenuated the association between sRAGE and mortality to some extent, although the association remained significant. Finally, further adjustment for relevant health indicators (models 3 and 4) showed that these had minimal effects. Altogether, the fully adjusted model suggest that frail individuals are three times more likely to die at any given time during a four year follow-up period per unit increment of the natural log- transformed sRAGE (HR=3.07, 95%CI 1 .32-7.14, P <0.009).

To evaluate for the potential effect of time of follow-up on the association between sRAGE and mortality the fully adjusted Cox regression analysis (equivalent to model 4 above) was repeated including follow-up mortality data for 5 and 6 years. As shown in Table 5 the estimated HR of death per unit increment of natural log-transformed sRAGE was attenuated with increased follow-up duration, although it remained significant after 6 years. sRAGE as a prognostic marker of mortality in frail individuals - Results from multivariate Cox proportional hazards regression models for different cut-off levels of sRAGE

To our knowledge there is no clinically relevant threshold for the categorization of sRAGE concentrations. In the absence of such a threshold, the distribution of sRAGE concentrations measured in frail participants was divided by the median, by tertiles and by quartiles, in order to assess the extent to which different cut-off values of sRAGE would affect the risk of death. As shown in Table 6 frail participants whose measured levels of sRAGE were equal or above the median of the distribution (1294 pg/mL) exhibited a higher risk of dying than those with levels of sRAGE that fell below the median. Similarly, a graded increase in the risk of death was seen with ascending tertiles or ascending quartiles of sRAGE. Furthermore, these associations were not attenuated with additional adjustments for sociodemographic and health indicators. Thus for example, frail participants whose measured sRAGE levels were equal to or above the median were 4.5 times more likely to die than those with levels of sRAGE below the median (HR=4.53, 95%CI 1 .55-13.24, P <0.006). Similarly, participants in the highest quartile of sRAGE were over 6-times more likely to die than those in the lowest quartile of sRAGE (HR=6.25, 95%CI 1.32-29.55, P <0.006).

The fully-adjusted hazard ratios of death for different cut-off values of sRAGE, estimated based on 4, 5 and 6 years follow-up data are presented in Table 7. These results show that the associations were more pronounced when restricting the analysis to the first 4 years of follow-up. sRAGE as a prognostic factor of mortality in frail individuals Results from multivariate Cox proportional hazards regression models including eGFR instead of creatinine as a covariate

Kidney impairment is known to affect sRAGE levels [16] Hence, to explore further the potential influence of renal function on the relationship between sRAGE and mortality, a stepwise multivariate Cox regression analysis which included eGFR instead of creatinine as a covariate was performed. As shown in Table 8, the results of this analysis, which in this case was carried out using the six years follow-up mortality data, were in keeping with those which were obtained based on the follow-up period of four years presented in Table 4. Namely, they showed that the association between sRAGE and mortality persisted across all models, both for the whole sample and for the frail group, such a relationship not being found in non-frail group. Similarly, like in the four years analysis, the estimated HR of death conferred by sRAGE was always larger in the frail group than in all the participants. In this instance, the fully adjusted model (model 4), including eGFR as one of the covariates, indicated that frail individuals are 2.7-fold more likely to die at any given time during the six year follow- up period per unit increment of natural log-transformed sRAGE (HR=2.72, 95%CI 1 .48-4.99, P=0.001 ). These results compare favourably to those corresponding to the six years follow-up analysis shown in Table 5 for the fully adjusted model, which included creatinine as a covariate instead, providing confirmation that renal function was not a confounding factor. We also explored whether there was an interaction between eGFR and sRAGE on mortality and found the interaction term to be non- significant. In contrast, there was a significant interaction between sRAGE and frailty, which was maintained across all the models (Table 8), providing additional evidence that frailty influences the association between sRAGE and mortality.

A secondary analysis based on six years follow-up data, and including eGFR as a covariate, where sRAGE levels in the frail subsample were divided by quartiles (see Table 9), showed that frail participants in the highest quartile of sRAGE had a 3.5-fold greater risk of death than those in the lowest quartile (HR=3.51 , 95%CI 1 .38-8.91 , P= 0.008). sRAGE as a prognostic marker of mortality in frail individuals - Results from multivariate Cox proportional hazards regression analysis for sRAGE quartiles of the analytical sample

Table 10 shows a quartile-based comparison of the relationship between sRAGE and mortality (based on six year follow-up data), with sRAGE categorised according to cut- offs set from the entire analytical sample. Consistent with the results obtained using the frail subsample quartiles, this alternative analysis showed that frail participants who fell above the highest cut-off level of sRAGE had a 2.6-fold greater risk of death than those that fell below the lowest cut-off (HR=2.62, 95%CI 1 .07-6.42, P=0.036). In contrast, no significant increase in the risk of death was observed using the same sRAGE cut-offs in the non-frail group (HR=1 .37, 95%CI 0.56-3.35, P=0.497). Kaplan Meyer survival analysis of non-frail and frail individuals for different cutoff levels of sRAGE

Figure 3 shows survival curves for the frail participants in the analytical sample according to different cut-off levels of sRAGE. The top panel shows that participants with levels of sRAGE above the median of 1294 pg/mL had lower survival rates compared to individuals with levels under the median (P=0.010, Cox-Mantel log rank test). The middle panel shows that individuals in the highest tertile (T3), with levels of sRAGE above 1626 pg/mL, had poorer survival rates compared to individuals in the two lower tertiles (P=0.015, log rank-test). The bottom panel shows that individuals in the highest quartile (Q4), with levels of sRAGE above 1887 pg/mL, had lower survival rates compared to individuals in the three lower quartiles (P=0.002, log rank test). Figure 4 shows the survival curves for the non-frail participants. The results clearly demonstrate that unlike in the case of frail-participants, in those that were not frail the survival curves for different cut-off levels of sRAGE were similar.

In summary the survival curves clearly demonstrate that in frail individuals the incidence of death was higher if the measured sRAGE level fell in a higher category, whether higher than the median, the highest tertile or the highest quartile of sRAGE, had a higher incidence of deaths than the lower sRAGE groups; these differences were not observed in non-frail individuals.

Area under the curve analysis (AUC) with and without sRAGE

Figure 5 shows the integrated area under the curve (IAUC) values calculated over the time horizon of the study for hazard models which included or excluded sRAGE. This analysis demonstrated stark differences between frail and non-frail individuals. In the frail group inclusion of sRAGE in the hazard model consistently increased its predictive accuracy at all times points. In contrast, for the non-frail group, inclusion of sRAGE showed no added discriminatory value over that afforded by using only traditional sociodemographic and health indicators. Thus, taken together these results suggest that sRAGE could improve the ability to correctly assess the risk of death in frail elderly individuals.

Summary We investigated the relationship between sRAGE and mortality in a prospective study of elderly subjects living in the community. Strikingly elevated total serum sRAGE is a marker of poor survival in frail older adults.

References

1 . Fried, L.P., et al. , Frailty in older adults: evidence for a phenotype. J Gerontol A Biol Sci Med Sci, 2001 . 56(3): p. M146-56.

2. Clegg, A., et al., Frailty in elderly people. Lancet, 2013. 381 (9868): p. 752-62.

3. Hudson, B. l. and M.E. Lippman, Targeting RAGE Signaling in Inflammatory Disease. Annu Rev Med, 2018. 69: p. 349-64.

4. Schmidt, A.M., Soluble RAGEs - Prospects for treating & tracking metabolic and inflammatory disease. Vascul Pharmacol, 2015. 72: p. 1 -8.

5. Litwinoff, E., et al., Emerging Targets for Therapeutic Development in Diabetes and Its Complications: The RAGE Signaling Pathway. Clin Pharmacol Ther, 2015. 98(2): p. 135-44.

6. Garcia-Garcia, F.J., et al., The prevalence of frailty syndrome in an older population from Spain. The Toledo Study for Healthy Aging. J Nutr Health Aging, 201 1 . 15(10): p. 852-6.

7. Peres, K., et al., Health and aging in elderly farmers: the AMI cohort. BMC. Public Health, 2012. 12: p. 558.

8. Pilleron, S., et al., Patterns of circulating fat-soluble vitamins and carotenoids and risk of frailty. Eur J Nutr, 2018, DOI: 10.1007/s00394-017-1602-0

9. Guadalupe-Grau, A., et al., Association of regional muscle strength with mortality and hospitalisation in older people. Age Ageing, 2015. 44(5): p. 790-5.

10. Ottenbacher, K.J., et al., The reliability of upper- and lower-extremity strength testing in a community survey of older adults. Arch Phys Med Rehabil, 2002. 83(10): p. 1423-7.

1 1 . Avila-Funes, J.A., et al., Is frailty a prodromal stage of vascular dementia? Results from the Three-City Study. J Am Geriatr Soc, 2012. 60(9): p. 1708-12.

12. Lyness, J.M., et al., Screening for depression in elderly primary care patients. A comparison of the Center for Epidemiologic Studies-Depression Scale and the Geriatric Depression Scale. Arch Intern Med, 1997. 157(4): p. 449-54. 13. Schuit, A.J., et al. , Validity of the Physical Activity Scale for the Elderly (PASE): according to energy expenditure assessed by the doubly labeled water method. J Clin Epidemiol, 1997. 50(5): p. 541-6.

14. Katz, S., et al., Studies of Illness in the Aged. The Index of Adi: A Standardized Measure of Biological and Psychosocial Function. JAMA, 1963. 185: p. 914-9.

15. Levey, A.S., et al., A new equation to estimate glomerular filtration rate. Ann Intern Med, 2009. 150: p. 604-12.

16. Colhoun H.M., et al. Total soluble and endogenous secretory receptor for advanced glycation end products as predictive biomarkers of coronary heart disease risk in patients with type 2 diabetes: an analysis from the CARDS trial. Diabetes, 2011. 60: p. 2379-85.

17. Tam, X.H., et al., Enhanced expression of receptor for advanced glycation end- products is associated with low circulating soluble isoforms of the receptor in Type 2 diabetes. Clin Sci, 2011. 120: p. 81 -9.

18. Zhang, L., et al., Receptor for Advanced Glycation End Products Is Subjected to Protein Ectodomain Shedding by Metalloproteinases. J Biol Chem, 2008. 283:, 35507-16.

Table 1 .

All Non-frail Frail

P

(n=691 ) (n=550) (n=141 )

Sociodemographic and behavioural parameters

Cohort, %

AMI 41.4 42.7 35.5 0.116

TSHA 58.6 57.3 64.5

Male, % 49.6 52.9 37.6 0.001

Age in years, mean ± SD 75.2 ± 6.0 74.1 ± 5.4 79.8 ± 6.1 <0.001

Education, %

Low 87.8 86.2 94.3

0.015

Intermediate 9.8 1 1.1 5.0

High 2.3 2.7 0.7

Smoking history, % 32.9 32.9 33.3 0.924

Laboratory and biomedical parameters

sRAGE, median [IQR], pg/mL 121 1 [923-1609] 1200 [923-1551] 1294 [916-1887] 0.094 Total cholesterol, mean ± SD, mg/dl_ 199.3 ± 40.0 200.1 ± 39.6 195.9 ± 41.5 0.280 Creatinine, median [IQR], mg/dl_ 0.9 [0.7-1.0] 0.9 [0.7-1.0] 0.9 [0.7-1.1] 0.160

BMI, median [IQR], kg/m ² 28.2 [25.5-31.2] 27.9 [25.4-30.8] 29.3 [25.8-32.9] 0.004 Obesity, % 34.9 32.4 44.4 0.010

ADL, % dependent 10.9 2.5 43.3 <0.001 Frailty, % 20.4

Diabetes, % 17.7 16.0 24.8 0.018

Hypertension, % 61.1 60.5 63.8 0.475

Cardiovascular disease, % 16.6 14.0 27.0 <0.001

Stroke, % 5.5 4.0 1 1.3 0.001

Cancer, % 8.7 8.9 8.5 0.882

Table 3.

All (n=691) Non-frail (n=550) Frail (n=141)

Variable HR 95% Cl P HR 95% Cl P HR 95% Cl P

Cohort 0.99 0.60-1.64 0.981 1.02 0.54-1.95 0.945 0.80 0.36-1.79 0.589

Age 1.14 1.10-1.19 <0.001 1.17 1.11-1.24 <0.001 1.08 1.02-1.16 0.011

Male gender 2.28 1.34-3.87 0.002 2.27 1.12-4.57 0.022 3.42 1.51-7.75 0.003 Education 0.61 0.24-1.52 0.288 0.73 0.26-2.06 0.556 0.63 0.08-4.64 0.649 Smoking history 2.05 1.25-3.36 0.004 1.88 1.00-3.56 0.052 2.42 1.10-5.31 0.027 Ln sRAGE 2.76 1.61-4.73 <0.001 1.55 0.75-3.23 0.238 4.18 1.91-9.16 <0.001 Total cholesterol 0.99 0.99-1.00 0.026 0.99 0.98-1.00 0.019 1.00 0.99-1.01 0.681 Creatinine 2.92 2.16-3.94 <0.001 2.70 1.74-4.17 <0.001 2.88 1.81-4.56 <0.001 BMI 0.96 0.91-1.01 0.138 0.97 0.90-1.04 0.396 0.93 0.86-1.00 0.060

Obesity 0.87 0.51-1.47 0.599 0.85 0.42-1.71 0.640 0.69 0.30-1.56 0.372

ADL 4.71 2.79-7.96 <0.001 5.43 1.93-15.31 0.001 3.04 1.31-7.04 0.010

Frailty 2.76 1.67-4.57 <0.001

Diabetes 2.48 1.47-4.19 0.001 1.68 0.80-3.55 0.173 3.25 1.48-7.12 0.003

Hypertension 1.38 0.81-2.35 0.230 1.61 0.80-3.25 0.181 1.05 0.46-2.38 0.904

Cardiovascular

3.11 1.86-5.20 <0.001 2.65 1.31-5.33 0.007 2.74 1.25-6.01 0.012 disease

Stroke 3.04 1.50-6.15 0.002 4.02 1.57-10.29 0.004 1.44 0.50-4.20 0.502

Cancer 1.10 0.48-2.56 0.818 1.20 0.42-3.37 0.734 1.05 0.25-4.46 0.947

Table 4.

All (n=691) Non-frail (n=550) Frail (n=141)

HR 95% Cl P HR 95% Cl P HR 95% Cl P

Unadjusted 2.76 1.61 -4.73 <0.001 1.55 0.75-3.23 0.238 4.18 1.91 -9.16 <0.001 Model 1 2.78 1.61 -4.77 <0.001 1.55 0.75-3.23 0.239 4.16 1.90-9.12 <0.001 Model 2 2.23 1.30-3.83 0.004 1.33 0.62-2.86 0.458 3.34 1.61 -6.93 0.001 Model 3 1.90 1.11 -3.27 0.020 1.48 0.65-3.37 0.356 3.02 1.30-7.06 0.010 Model 4 1.89 1.10-3.26 0.022 1.44 0.62-3.31 0.392 3.07 1.32-7.14 0.009

Table 5

All (n=691) Non-frail (n=550) Frail (n=141)

Period HR 95% Cl P HR 95% Cl P HR 95% Cl P

4-years 1.89 1.10-3.26 0.022 1.44 0.62-3.31 0.392 3.07 1.32-7.14 0.009

5-years 1.80 1.12-2.91 0.016 1.39 0.65-2.99 0.394 2.36 1.23-4.51 0.010

6-years 1.76 1.13-2.74 0.013 1.47 0.72-2.97 0.287 2.39 1.28-4.44 0.006

N>

Table 6

Model 1 Model 2 Model 3 Model 4

HR 95% Cl P HR 95% Cl P HR 95% Cl P HR 95% Cl P sRAGE median

M1 , <1294pg/ml_ 1 (ref) 1 (ref) 1 (ref) 1 (ref)

M2, ³1294pg/ml_ 3.39 1.35-8.49 0.009 3.51 1.40-8.84 0.008 3.98 1.42-11.16 0.009 4.53 1.55-13.24 0.006 sRAGE tertiles

T1 , <1035pg/mL 1 (ref) 1 (ref) 1 (ref) 1 (ref)

T2, 1035-1626pg/ml_ 1.21 0.37-3.96 0.755 1.07 0.32-3.58 0.917 1.88 0.52-6.83 0.336 1.86 0.52-6.73 0.342 T3, >1626pg/mL 3.01 1.08-8.37 0.035 3.32 1.17-9.40 0.024 3.71 1.10-12.55 0.035 3.82 1.12-12.95 0.032 sRAGE quartiles

Q1 , <915.7pg/mL 1 (ref) 1 (ref) 1 (ref) 1 (ref)

Q2, 915.7-1293.9pg/mL 1.04 0.21-5.15 0.963 0.92 0.18-4.73 0.925 1.52 0.23-9.90 0.663 1.43 0.22-9.08 0.706

Q3, 1294-1887pg/mL 2.06 0.52-8.25 0.306 2.01 0.49-8.20 0.329 3.71 0.78-17.69 0.100 4.22 0.85-20.87 0.078

Q4, >1887pg/mL 5.03 1.43-17.67 0.012 4.97 1.38-17.97 0.014 5.95 1.26-28.03 0.024 6.25 1.32-29.55 0.021

Table 7

4-years 5years 6 years

HR 95% Cl P HR 95% Cl P HR 95% Cl P sRAGE median

M1 , <1294 pg/mL 1 (ref) 1 (ref) 1 (ref)

M2, ³1294 pg/mL 4.53 1.55-13.24 0.006 3.12 1.50-6.50 0.002 2.54 1.32-4.88 0.005 sRAGE tertiles

T1 , <1035 pg/mL 1 (ref) 1 (ref) 1 (ref)

-i^. T2, 1035-1626 pg/mL 1.86 0.52-6.73 0.342 1.48 0.61-3.60 0.391 1.19 0.54-2.65 0.668

T3, >1626 pg/mL 3.82 1.12-12.95 0.032 2.81 1.19-6.67 0.019 2.40 1.12-5.14 0.024 sRAGE quartiles

Q1 , <915.7 pg/mL 1 (ref) 1 (ref) 1 (ref)

Q2, 915.7-1293.9 pg/mL 1.43 0.22- 9.08 0.706 0.86 0.25-2.96 0.807 0.62 0.20-1.91 0.406 Q3, 1294-1887 pg/mL 4.22 0.85-20.87 0.078 2.22 0.76-6.54 0.147 1.36 0.54-3.45 0.512 Q4, >1887 pg/mL 6.25 1.32-29.55 0.021 3.45 1.19-10.00 0.022 2.82 1.10-7.25 0.032

Table 8.

All (n=691) Non-frail (n=550) Frail (n=141)

Unadjusted 2.20 1.43-3.36 <0.001 1.51 0.81-2.82 0.196 2.69 1.53-4.76 0.001 <0.001 Model 1 2.25 1.46-3.45 <0.001 1.53 0.82-2.87 0.185 2.65 1.51-4.66 0.001 0.001 Model 2 1.81 1.17-2.80 0.008 1.34 0.70-2.58 0.379 2.30 1.33-3.98 0.003 0.007 Model 3 1.84 1.19-2.84 0.006 1.45 0.73-2.91 0.289 2.73 1.48-5.03 0.001 0.005 Model 4 1.85 1.19-2.86 0.006 1.43 0.71-2.89 0.312 2.72 1.48-4.99 0.001 0.004

P value for the interaction between sRAGE and frailty

Table 9.

Model 1 Model 2 Model 3 Model 4

sRAGE quartiles

Q1 , <915.7 pg/mL 1 (ref) 1 (ref) 1 (ref) 1 (ref)

Q2, 915.7-1293.9 pg/mL 0.86 0.34-2.18 0.746 0.61 0.23-1.61 0.318 0.77 0.26-2.35 0.651 0.73 0.24-2.23 0.584 Q3, 1294-1887 pg/mL 1.26 0.54-2.92 0.589 1.06 0.44-2.52 0.903 1.35 0.54-3.39 0.522 1.41 0.56-3.58 0.471 Q4, >1887 pg/mL 2.87 1.34-6.15 0.007 2.53 1.15-5.59 0.021 3.61 1.43-9.12 0.007 3.51 1.38-8.91 0.008 o\

Table 10.

All (n=691) Non-frail (n=550) Frail (n=141)

sRAGE quartiles*

Q1 , <923.4 pg/mL 173 1 (ref) 138 1 (ref) 35 1 (ref)

Q2, 923.4-1211.1 pg/mL 173 0.70 0.36-1.38 0.307 144 1.10 0.45-2.69 0.832 29 0.76 0.24-2.48 0.653 Q3, 1211.2-1608.7 pg/mL 172 1.27 0.70-2.31 0.438 143 1.68 0.74-3.86 0.218 29 1.11 0.42-2.94 0.838 Q4, >1608.7 pg/mL 173 1.63 0.92-2.91 0.097 125 1.37 0.56-3.35 0.497 48 2.62 1.07-6.42 0.036

^* Cut-offs based on quartiles of the entire analytical sample