Elisa Heikkilä D 1957 AN N ALES UN IVERSITATIS TURKUEN SIS TURUN YLIOPISTON JULKAISUJA – ANNALES UNIVERSITATIS TURKUENSIS SARJA – SER. D OSA – TOM. 1957 | MEDICA – ODONTOLOGICA | TURKU 2026 How Can Laboratory Parameters Aid Clinicians in Prediction of Cardiovascular Events, Institutionalization and Mortality in Older People? Elisa Heikkilä Elisa Heikkilä HOW CAN LABORATORY PARAMETERS AID CLINICIANS IN PREDICTION OF CARDIOVASCULAR EVENTS, INSTITUTIONALIZATION AND MORTALITY IN OLDER PEOPLE? TURUN YLIOPISTON JULKAISUJA – ANNALES UNIVERSITATIS TURKUENSIS SARJA – SER. D OSA – TOM. 1957 | MEDICA – ODONTOLOGICA | TURKU 2026 University of Turku Faculty of Medicine Department of Clinical Medicine Clinical Chemistry Doctoral Programme in Clinical Research Supervised by Adjunct professor Marika Salminen Department of General Practice University of Turku Turku, Finland Adjunct professor Laura Viikari Department of Geriatric Medicine, University of Turku and Turku University Hospital Turku, Finland Professor emeritus Kari Pulkki Faculty of Medicine University of Helsinki and Diagnostic Center, Helsinki University Hospital Helsinki, Finland Reviewed by Associate professor Hanna-Maria Roitto Faculty of Medicine and Health Technology University of Tampere Tampere, Finland Professor Tiina Laatikainen Institute of Public Health and Clinical Nutrition Faculty of Medicine University of Eastern Finland Kuopio, Finland Opponent Professor Tuija Männistö Faculty of Health Sciences University of Eastern Finland Kuopio, Finland The originality of this publication has been checked in accordance with the University of Turku quality assurance system using the Turnitin OriginalityCheck service. Cover image: Otava Heikkilä ISBN 978-952-02-0590-4 (PRINT) ISBN 978-952-02-0591-1 (PDF) ISSN 0355-9483 (Print) ISSN 2343-3213 (Online) Painosalama, Turku, Finland 2026 Jussille UNIVERSITY OF TURKU Faculty of Medicine Clinical Chemistry ELISA HEIKKILÄ: How can laboratory parameters aid clinicians in prediction of cardiovascular events, institutionalization and mortality in older people? Doctoral Dissertation, 125 pp. Doctoral Program in Clinical Research March 2026 ABSTRACT Background: Cardiovascular diseases are the most common cause of death worldwide, and a major cause for disability and mortality in older people. Laboratory parameters can be of aid in predicting cardiovascular events and mortality. Indexes that combine different parameters can help in risk prediction. Many previous indexes include a large number of clinical and/or laboratory parameters, which complicates their use in clinical practice. Aims: The aim of the thesis was to study how a combination of routine laboratory parameters can be used to estimate an older individual’s risk for institutionalization and mortality. Also, the aim was to study how cardiac markers troponin T (cTnT) and N-terminal pro-brain natriuretic peptide (proBNP) can estimate both cardiovascular outcomes and all-cause mortality, and how the concentrations of these markers change with ageing. Results: Older people’s mortality could be predicted with fourteen routine laboratory parameters, or a combination of six clinical and three laboratory parameters. Laboratory parameters could not predict institutionalization, but it could be predicted with three clinical parameters. Elevated concentrations of cTnT and proBNP both predict cardiac events, and cardiovascular and all-cause mortality, but their concentrations rise with normal ageing as well. Conclusions: Only a small number of routine laboratory parameters as well as cTnT and proBNP above age-adjusted reference limits help to predict mortality in older people. On the contrary, institutionalization is associated with clinical parameters which reflect difficulties of activities in daily living. KEYWORDS: laboratory tests, cardiac markers, mortality, institutionalization, cardiac events, older people 5 TURUN YLIOPISTO Lääketieteellinen tiedekunta Kliininen kemia ELISA HEIKKILÄ: Miten laboratoriokokeet auttavat lääkäriä sydän- ja verisuonitapahtumien, pitkäaikaishoidon tarpeen ja kuolleisuuden ennustamisessa? Väitöskirja, 125 s. Turun kliininen tohtoriohjelma Maaliskuu 2026 TIIVISTELMÄ Tausta: Sydän- ja verisuonisairaudet ovat yleisin kuolinsyiden ryhmä maailmanlaa- juisesti, sekä merkittävä ikääntyneiden toimintakyvyn laskun ja kuolleisuuden aihe- uttaja. Laboratoriokokeilla voidaan ennustaa sydän- ja verisuonitautitapahtumia ja kuolleisuutta. Eri parametreja yhdistävät indeksit voivat auttaa riskinarvioinnissa. Monet käytetyt indeksit sisältävät hyvin suuren määrän kliinisiä ja/tai laboratorio- parametreja, mikä tekee niiden käytöstä kliinisessä työssä hankalaa. Tavoitteet: Väitöskirjatyön tavoitteena oli tutkia, kuinka rutiininomaiset laboratorio- kokeet voivat auttaa arvioitaessa ikääntyneen riskiä päätyä pitkäaikaishoitoon tai kuolla. Lisäksi tavoitteena oli selvittää, kuinka sydänmerkkiaineet troponiini T (cTnT) ja b-tyypin N-terminaalinen natriureettinen propeptidi (proBNP) ennustavat sekä sydän- ja verisuonitapahtumia että kuolleisuutta, sekä kuinka näiden merkki- aineiden pitoisuudet muuttuvat ikääntymisen myötä. Tulokset: Ikääntyneiden kuolleisuutta voitiin ennustaa neljällätoista rutiinikäytössä olevalla laboratorioparametrilla, tai kuuden kliinisen ja kolmen laboratoriopara- metrin yhdistelmällä. Laboratorioparametreilla ei voitu ennustaa pitkäaikaishoitoon päätymistä, mutta sitä voitiin ennustaa kolmella kliinisellä parametrilla. cTnT:n ja proBNP:n kohonneet pitoisuudet ennustivat sydän- ja verisuonitapahtumia sekä kuolleisuutta niin sydänsairauksiin kuin muihinkin syihin, mutta niiden pitoisuudet nousivat myös normaalin ikääntymisen myötä. Johtopäätökset: Pieni määrä rutiinikäytössä olevia laboratoriokokeita samoin kuin iän mukaiset viitearvot ylittävät cTnT- ja proBNP –pitoisuudet auttavat ikäänty- neiden kuolleisuuden ennustamisessa. Sen sijaan pitkäaikaishoitoon päätyminen on yhteydessä kliinisiin parametreihin, jotka heijastavat päivittäistoiminnoista selviyty- mistä. AVAINSANAT: laboratoriokokeet, sydänmerkkiaineet, kuolleisuus, pitkäaikais- hoito, sydäntapahtumat, ikääntyneet 6 Table of Contents Abbreviations .................................................................................. 8 List of Original Publications ........................................................... 9 1 Introduction ........................................................................... 10 2 Review of the Literature ....................................................... 12 2.1 Ageing population .................................................................. 12 2.2 Cardiovascular ageing ........................................................... 15 2.3 Risk factors for cardiovascular disease .................................. 16 2.3.1 Conventional risk factors ............................................. 16 2.3.2 Non-traditional markers for cardiovascular risk ............ 17 2.3.3 Cardiovascular risk factors in old age .......................... 18 2.4 Reference intervals ................................................................ 20 2.4.1 Defining reference intervals ......................................... 20 2.4.2 Older people’s reference intervals ............................... 20 2.5 Cardiac markers ..................................................................... 21 2.5.1 Cardiac troponins ........................................................ 21 2.5.2 N-terminal b-type natriuretic propeptide ...................... 22 2.6 Frailty ..................................................................................... 23 2.7 Frailty indexes ........................................................................ 24 2.7.1 Clinical indexes ........................................................... 25 2.7.2 Laboratory indexes ...................................................... 26 2.7.3 Combined indexes ...................................................... 27 2.8 Institutionalization................................................................... 28 2.9 Mortality ................................................................................. 29 3 Aims ....................................................................................... 30 4 Material and Methods ........................................................... 31 4.1 Study population .................................................................... 31 4.2 Methods ................................................................................. 32 4.2.1 Baseline examination .................................................. 32 4.2.2 Laboratory measurements .......................................... 32 4.2.3 Institutionalization ........................................................ 33 4.2.4 Mortality ...................................................................... 33 4.2.5 The incidence of acute myocardial infarctions ............. 33 4.2.6 The laboratory index ................................................... 33 4.2.7 The combined clinical and laboratory index ................. 34 4.3 Statistical analyses................................................................. 36 7 4.4 Ethics ..................................................................................... 37 5 Results ................................................................................... 38 5.1 Characteristics of the participants ........................................... 38 5.2 Index studies .......................................................................... 40 5.2.1 The laboratory index .................................................... 40 5.2.2 The combined index .................................................... 42 5.2.3 The reduced indexes ................................................... 43 5.3 Cardiac marker studies........................................................... 48 5.3.1 Baseline characteristics of the reference population (Study III)..................................................................... 48 5.3.2 Reference limits for troponin T and N-terminal b- type natriuretic propeptide (Study III) ........................... 49 5.3.3 The associations of troponin T and N-terminal b- type natriuretic propeptide with the ten-year incidence of acute myocardial infarctions (Study III) .... 51 5.3.4 The associations of troponin T and N-terminal b- type natriuretic propeptide with cardiac and all- cause mortality (Study IV) ........................................... 51 6 Discussion ............................................................................. 53 6.1 Study population..................................................................... 54 6.2 Findings ................................................................................. 55 6.2.1 Laboratory index (Study I) ........................................... 55 6.2.2 Combined indexes (Study II) ....................................... 58 6.2.3 Reference limits for troponin T and N-terminal b-type natriuretic propeptide in older population (Study III) ..... 60 6.2.4 Prediction of cardiovascular events and cardiovascular and all-cause mortality with troponin T and N-terminal b-type natriuretic propeptide (Studies III and IV) ....................................................... 62 6.3 Strengths and limitations ........................................................ 63 6.4 Implications for clinics and future research ............................. 66 6.4.1 Implications for clinics.................................................. 66 6.4.2 Implications for future research ................................... 68 7 Conclusions ........................................................................... 70 Acknowledgements ....................................................................... 71 References ..................................................................................... 73 List of Figures and Tables ............................................................ 88 Original Publications ..................................................................... 91 8 Abbreviations ALP Alkaline phosphatase ALT Alanine aminotransferase AMI Acute myocardial infarction AUC Area under the curve CI Confidence interval CRP C-reactive protein cTnT Cardiac troponin T CV Cardiovascular CVD Cardiovascular disease ECLIA Electrochemiluminescence immunoassay FI Frailty index HDL High-density lipoprotein HR Hazard ratio ICD-10 10th Revision of International Statistical Classification of Diseases and Related Health Problems IQR Interquartile range LDL Low-density lipoprotein LI Laboratory index LoD Limit of detection LoQ Limit of quantitation NPV Negative predictive value PPV Positive predictive value proBNP N-terminal natriuretic b-type propeptide SD Standard deviation TSH Thyroid stimulating hormone 9 List of Original Publications This dissertation is based on the following original publications, which are referred to in the text by their Roman numerals: I Heikkilä E, Salminen M, Viljanen A, Katajamäki T, Koivula MK, Pulkki K, Isoaho R, Kivelä SL, Viitanen M, Löppönen M, Vahlberg T, Viikari L, Irjala K. A practical laboratory index to predict institutionalization and mortality - an 18-year population-based follow-up study. BMC Geriatrics 2021 Feb 25;21(1):139. doi: 10.1186/s12877-021-02077-1. II Heikkilä E, Salminen M, Viljanen A, Katajamäki T, Koivula MK, Pulkki K, Isoaho R, Kivelä SL, Viitanen M, Löppönen M, Vahlberg T, Venäläinen MS, Elo LL, Viikari L, Irjala K. A novel easy-to-use index to predict institutionalization and death in older population - a 10-year population-based follow-up study. BMC Geriatrics 2023 Feb 7;23(1):80. doi: 10.1186/s12877- 023-03760-1. III Heikkilä E, Katajamäki T, Salminen M, Irjala K, Viljanen A, Koivula MK, Pulkki K, Isoaho R, Kivelä SL, Viitanen M, Löppönen M, Vahlberg T, Viikari L. New reference limits for cardiac troponin T and N-terminal b-type natriuretic propeptide in elders. Clinica Chimica Acta 2024 Mar 15;556:117844. doi: 10.1016/j.cca.2024.117844. IV Heikkilä E, Katajamäki T, Salminen M, Irjala K, Viljanen A, Koivula MK, Pulkki K, Viitanen M, Vahlberg T, Viikari L. High-sensitivity cardiac troponin T and N-terminal b-type natriuretic propeptide are associated with cardiac and all-cause mortality in older adults - A population-based ten-year follow-up study. Clinica Chimica Acta 2025 Feb 1;567:120116. doi: 10.1016/j.cca.2024.120116. The original publications have been reproduced with the permission of the copyright holders. 10 1 Introduction The average lifespan is increasing. The world’s population is expected to continue to grow over the coming decades, until reaching a peak of over 10 billion people in the 2080s. By then, the number of over 65-year-olds is estimated to surpass the number of under 18-year-olds, being over 2 billion (United Nations, 2024). Developments in medical treatments and new medicines and advances in medical technology and improvements in the access to health care have contributed in global population ageing and at the same time increased the prevalence of chronic diseases. The challenge of ageing research is to find ways to increase healthy lifespan. While the average lifespan has increased, the percentual time lived with disability has not decreased. Thus, the absolute number of years lived with disability has even grown and was estimated to be nine years in 2020 (World Health Organization, 2024). The aim should be to compress morbidity to less years lived with disability (Garmany et al., 2021). With more disease-free and functional additional years, it could be possible to improve survival at home and the quality of life in later years. Persons at the same chronological age can have very different functional and cognitive capacities (Mitnitski et al. 2002). On average, organ system functions decline with age, but the rate varies, within and between individuals, also in people who are otherwise well. There is a tendency towards more personalized health care, where treatment could be more optimized and targeted to the individual. Multiple factors contribute to the individual rate of ageing some of which can be modified by medical interventions or personal habits of life. Much is already known about healthy life habits that prevent declining of physical condition such as healthy nutrition and adequate physical exercise. Also, social and mental factors are known to contribute to healthy ageing. One way to estimate a person’s state of health and susceptibility to adverse health risks is to calculate an index score that combines clinical or laboratory defects that an individual possesses. This kind of an index can be called a frailty index (FI). Often these indexes contain a large number such as over thirty clinical and/or laboratory parameters. Various FIs have been used to predict older people’s health risks and mortality. It has been proposed that a person’s biological age could be estimated by comparing their FI score with the average FI score at their age. As frailty predicts adverse effects better than Introduction 11 chronological age, it may serve as a proxy measure of biological age (Mitnitski et al., 2001, 2002; Clegg et al., 2013). Ageing is the largest risk factor for cardiovascular disease (CVD) (North & Sinclair, 2012). On the other hand, maintaining cardiovascular (CV) health is essential for the whole organism. Normal circulatory function is an important factor in determining the whole organism’s longevity. Throughout lifespan, the CV system is under constant mechanical and metabolic stress (Abdellatif et al., 2023). Nevertheless, while many people acquire CV and other diseases by old age, there are also many who achieve old age without evidence of these diseases. There are typical age-related changes seen in cardiac function, but some still preserve good performance in the old age (Lakatta, 2002). People have often developed many comorbidities and risk factors for CVD by old age. The prevalence of pathologies that affect the CV system is still growing, and they are the leading cause of global morbidity, disability and mortality (Abdellatif et al., 2023), resulting in 40 percent of deaths in over 65-year-olds, and the cost associated with treatment is also still increasing (Fleg et al., 2011; North & Sinclair, 2012). Prevention and early detection of CVD is crucial so that treatment, including counselling and medication, can start promptly. There is evidence on the benefits of treating CV risk factors also in older age (Bulló et al., 2011; Yang et al., 2021; Lam et al., 2007). CVD risk is traditionally assessed with models that include known risk factors for CVD: high blood pressure, cigarette smoking, dyslipidemia, obesity and diabetes, some also include the elevated risk with higher age and male gender. Many models include some laboratory values such as cholesterol levels. The role of the risk factors may change in older age, and the traditional risk evaluation methods may not perform well especially in primary prevention in older population. Measuring levels of CV biomarkers could improve the accuracy of risk stratification especially in individuals with an intermediate risk score and potentially lead to more intensive preventive strategies to those with elevated biomarker levels. The demand for more precise risk assessment methods continues to grow as newer, costly medications are introduced into clinical practice, and more diverse treatments for CVD prevention are available. Therefore, the focuses of this thesis were to study how the concentrations of cardiac biomarkers change with age, and how they could be used in risk prediction, and to study the possibilities to predict risks for overall mortality and institutionalization with a combination of a small number of parameters. 12 2 Review of the Literature 2.1 Ageing population Older population is a diverse group of individuals with different backgrounds, and in different states of health and physical condition. In addition to physical age, the biological age of older people varies. Older population can be defined as 65-year- olds and older (Organisation for Economic Co-operation and Development, n.d.). The proportion of over 65-year-olds is growing worldwide (United Nations, 2024), and in Finland (Figures 1, 2 and 3; Official Statistics Finland, 2024). While many over 65-year-olds are still in a good condition, it is noteworthy that also the number of over 75- and over 85-year-olds has grown and is estimated to continue to grow rapidly (Figure 3). The proportion of 65-year-olds and older in Finland was 7.3 in 1960, and has grown each year, being 22.6 in 2020, and 23.6 in 2024. Only 0.1 percent of the population in Finland were aged 85 years or over in 1950, and their percentage has grown to 3.0 percent in 2024 (Official Statistics Finland, 2025). Both the average expected life span has increased worldwide as well as healthy life expectancy. Life expectancy increased globally by more than 6 years between 2000 and 2019, being 73.1 years in 2019. Healthy life expectancy increased by 5.4 years during the same period, from 58.3 in 2000 to 63.7, in 2019. As the increase seen in healthy life expectancy is not quite as large as the increase in life expectancy (5.3 years vs 6.4 years), this means that there was no decrease in the years lived with disability. The prevalence of many diseases increases with advancing age. The COVID-19 pandemic reversed the increase in both life expectancy and healthy life expectancy, both retreating to 2012 levels in 2021 (71.4 years and 61.9 years, respectively). (World Health Organization, 2024.) The increasing older population, with no reduction in the years lived with disability, creates a grand challenge to health care systems worldwide. Considering the costs of institutionalization, there is a tendency and a necessity to reduce institutional care and enable old people to live at home longer with home care. This is problematic since the resources are scarce and not all old people in home care are in a healthy enough cognitive or physical condition to live alone. Review of the Literature 13 Figure 1. Age structure of population by age group and gender in Finland 31.12.2024 and 2075, projection 2024. Source: Official Statistics Finland [referred: 27.8.2025] Access method: https://stat.fi/en/statistics/vaenn. Reprinted according to the Creative Commons Attribution 4.0 International License. Figure 2. Population 1870–2024 and population projection 2025–2075 in Finland by age group, under 15-year-olds indicated by black line, 15 to 64-year-olds by blue line, and 65-year- olds and older by red line. Source: Official Statistics Finland [referred: 27.8.2025] Access method: https://stat.fi/en/statistics/vaenn. Reprinted according to the Creative Commons Attribution 4.0 International License. Fi gu re 3 . O ve r 6 5- ye ar -o ld p op ul at io n 20 07 –2 02 4 an d po pu la tio n pr oj ec tio n 20 25 –2 07 5 in F in la nd b y ag e gr ou ps 6 5– 74 , 7 5– 84 a nd 8 5– . C re at ed fr om da ta o f O ffi ci al S ta tis tic s Fi nl an d [R ef er en ce d: 4 .1 1. 20 25 ]. D at a ac ce ss m et ho d: h ttp s: //p xd at a. st at .fi P XW eb /p xw eb /e n/ St at Fi n/ . Elisa Heikkilä 14 Review of the Literature 15 2.2 Cardiovascular ageing Numerous physiological processes decline with age which leads to an increased risk of disease (Abdellatif et al., 2023). It is not always clear which symptoms or features are part of a normal ageing process, and which are due to a pathological condition. Ageing has a major effect on the heart and arterial systems leading to an increase in CVDs including atherosclerosis, hypertension, myocardial infarction, and stroke (Lakatta & Levy, 2003; North & Sinclair, 2012). There are different proposed mechanisms about why the prevalence of CVDs increases with advancing age. In addition to extrinsic stress, organisms face constant intrinsic stress such as reactive oxygen species produced by metabolism. Cell senescence leads to the loss of replicative capacity and deterioration of cell components and cellular functions. These factors inevitably cause accumulation of damage, leading to alterations in the homeostasis of tissues and eventually the whole organism (Vicencio et al., 2008). Also misfolded and damaged proteins and mutations accumulate with age (Abdellatif et al., 2023). There are multiple repair mechanisms responding to damage such as telomerase enzyme that repairs chromosomal damage. Telomerase activity seems to be one factor involved in determining longevity (Nicholls et al., 2011). Clearance mechanisms such as autophagy are important as well, and especially in terminally differentiated cells like cardiomyocytes (Kirkwood, 2011; Vicencio et al., 2008). Limitations in repair mechanisms lead to age-related pathologies and declining health over time (Vicencio et al., 2008). The heart undergoes characteristic changes also in individuals without a CVD. The still partly unknown genetic pathways that regulate ageing, also influence CV ageing (Lakatta & Levy, 2003; North & Sinclair, 2012). Mitochondrial function declines with age in most tissues, and this ageing affects also myocardial cells that are long-lived cells rich in mitochondria and a high demand of energy (North & Sinclair, 2012). The ageing process changes CV structure and function which enhances the risk for pathophysiological disease mechanisms. The complex interactions involve age, multiple risk factors, and genetics (Lakatta & Levy, 2003; Abdellatif et al., 2023). Some pathological alterations of the heart include hypertrophy, alterations in the left ventricular diastolic function and systolic capacity, increased arterial stiffness, and impairment of the endothelial function (Abdellatif et al., 2023; Lakatta & Levy, 2003). The health of the arterial and cardiac systems is linked. Myocardial hypertrophy and fibroblast proliferation can be compensatory mechanisms for increasing arterial stiffness. The number of cardiac myocytes decreases with age, which causes additional load due to stretching of the cells that remain. The increase in the size of the remaining myocytes and in the amount of collagen thickens the left ventricle wall (Gazoti Debessa et al., 2001). Also, the atrial size increases in otherwise healthy aged people, and the secretory behavior of the myocytes alters so that, for example, the secretion of Elisa Heikkilä 16 atrial natriuretic peptides is increased (Lakatta, 2002). These changes result in decreased cardiac output and increase in the amount of fibrotic tissue (Lakatta & Levy, 2003). Structural changes of the heart, including fibrosis and hypertrophy, slow propagation of electric impulse throughout the heart. The number of cells in the sinoatrial node decreases influencing heart rate. These age-related changes decrease both rate variability and maximum heart rate (Antelmi et al., 2004). All these changes reduce the compensatory capacity of the aged heart and make the heart more prone to myocardial dysfunction and failure (Olivetti et al., 1991). Cardiomyocytes are renewed at a very slow pace of approximately 1% per year at the age of 20, and 0.3% at the age of 75. Myocardial cells still have some regenerative potential as the rate of renewal of cardiomyocytes can increase after myocardial damage (Bergmann et al., 2009). While ageing increases the prevalence of CVDs, it has been speculated that CV ageing might precede or underlie age-related deterioration of overall health. The function of the CV system is vital in preserving the overall health as it provides blood circulation delivering oxygen and nutrients to all tissues, including the muscle and the brain (Abdellatif et al., 2023). 2.3 Risk factors for cardiovascular disease 2.3.1 Conventional risk factors The most important known traditional risk factors for developing CVD besides age and male gender include high blood pressure, cigarette smoking, dyslipidemia, obesity and diabetes, that account for approximately half of the global burden of CVD (The Global Cardiovascular Risk Consortium, 2025; Magnussen et al., 2023). Risk scores that combine these factors to assess an individual's likelihood of experiencing adverse CV outcomes have been developed; some considering additional factors such as medication use or geographical region. While age and gender are non-modifiable risk factors, they aid in understanding the overall risk. They are commonly included in risk-prediction models that are used in tailoring prevention (Conroy et al., 2003; D’Agostino et al., 2008; Wilson et al., 1998). Compared to risk scores designed for the general adult population, some differences are expected when estimating risks in older population. Conventional risk prediction models often do not consider the competing risk of death from non- CV causes, which may lead to overestimation of the benefits of risk factor treatment in older persons. A modified risk score for older people over the age of 70 has been developed as the relationship between traditional risk factors and CVD attenuates with age. (SCORE2 working group and ESC Cardiovascular risk collaboration, Review of the Literature 17 2021; SCORE2-OP working group and ESC Cardiovascular risk collaboration, 2021; Neumann et al., 2025.) 2.3.2 Non-traditional markers for cardiovascular risk The discriminatory value of most CV risk scores is moderate (Neumann et al., 2025). Emerging risk factors and biomarkers that are not part of traditional risk models but have been demonstrated to be independently linked to CV risk are gaining evidence for their ability to offer additional insights beyond traditional risk factors. Factors that have been studied are high-risk ethnicities, unfavourable body composition, presence of chronic inflammatory conditions, chronic kidney disease, genetic data, CV imaging results and circulating biomarkers (Neumann et al., 2025; Wang et al., 2023). Some of the biomarkers that could potentially enhance the accuracy of CV risk prediction reflect conditions that predispose the person to develop a CVD. C- reactive protein (CRP) and fibrinogen, markers for inflammation, and creatinine and cystatin C, markers for impaired kidney function, are associated with CVD (The Emerging Risk Factors Collaboration, 2012; Perone et al., 2025). Also, low and high alanine aminotransferase, high homocysteine levels and hyperuricemia have been found to be associated with increased CV mortality (Gospodarczyk et al., 2022; Moshkovits et al., 2024; Ndrepepa & Kastrati, 2019; Wang et al., 2023). Still, the incremental value of biomarkers when added to established risk models is modest. The identification of new biomarkers and optimal target populations may enhance their value in risk prediction in the future (Neumann et al., 2025). There are also newer biomarkers related to lipid metabolism. Lipoprotein (a) is a genetically determined lipoprotein variant with atherogenic and prothrombotic properties linked to the risk of CVD. European guidelines have suggested its measurement at least once for each person (Perone et al., 2025; Visseren et al., 2021). More recently, another class of lipids called ceramides has been recognized to predict CV end points. Risk scores based on ceramide levels could be used as efficient risk stratifiers in both primary and secondary prevention of atherosclerotic CVD, but their measurement requires specialized analytics (Katajamäki et al, 2022; Hilvo et al., 2020). CV biomarkers, including cardiac troponins and natriuretic peptides, can also be used to classify CV risk, especially in individuals at intermediate risk (Neumann et al., 2025). The role of troponins and natriuretic peptides in long-term risk prediction for CV events and mortality is discussed later in more detail. Elisa Heikkilä 18 2.3.3 Cardiovascular risk factors in old age Most CV risk factors, not including age and gender, can be controlled by addressing behavioral factors such as an unhealthy diet, obesity, and physical inactivity. In older adults, advancing age is the predominant CV risk factor. Older people are under- represented in primary and secondary prevention trials, and especially those who have comorbidities or are frail (Abrigani et al., 2024). There is some evidence on the benefits of addressing risk factors still in older age, although the role of risk factors differs from younger age groups (Bulló et al., 2011; Makino et al., 2021; Yang et al., 2021; Vilela de Sousa et al., 2023; Abrigani et al., 2024). Obesity may not be a similar risk factor in the older population as in younger population, but in younger old age (before the age of 70) moderate weight loss may still be beneficial in overweight older people to reduce their risk for diabetes, CVDs and mobility difficulties. In older population, the risk for total mortality caused by obesity is smaller than the risks caused by unintentional weight loss. (Strandberg et al., 2013; Suominen et al., 2014.) Type 2 diabetes mellitus significantly increases the risk of CVD. In high- income countries, individuals with diabetes face, on average, a twofold higher likelihood of developing CVD outcomes compared to those without diabetes (Pennells et al., 2023; Sarwar et al., 2010). Diabetes and hyperglycemia increase total and CV mortality. The earlier the person’s age at the onset of diabetes, the greater its effect on total and CV mortality (Abdelhafiz & Sinclair, 2022; Kaptoge et al., 2023; Morley et al., 2004). A study in an older population found metabolic syndrome strongly associated with CVD risk, and on the other hand higher weight was associated with the risk for metabolic syndrome. In this study, the older obese individuals without metabolic syndrome did not have a higher risk for CVD and concluded that targeting cardiometabolic risk factors could be considered in older people, regardless of weight status (Dhana et al., 2016). Many CV risk factors are closely connected. Obesity is associated with the risk for metabolic syndrome and lower physical activity, which are all risk factors for CVD (Dhana et al., 2016; Yang et al., 2021). A study on older men found light or vigorous physical activity to be strongly and negatively associated with CV as well as all-cause mortality. In the same study, both cessation of smoking and increasing physical activity even in old age reduced mortality (Holme & Anderssen, 2015). Cigarette smoke accelerates the development of atherosclerosis by enhancing oxidation of lipids, thrombosis, inflammation, arterial stiffness, and endothelial dysfunction (Ambrose & Barua, 2004). Smoking is a well-known risk factor for CVD and mortality in general population, and there is some evidence of the benefits of the cessation of smoking also in old age although the excess risk of smoking may be less apparent in the oldest old where smokers are increasingly strongly selected survivors (Lam et al., Review of the Literature 19 2007), and because the benefits might not be achieved during the first years after quitting smoking (Wei et al., 2021). Total cholesterol and low-density lipoprotein (LDL) cholesterol are risk factors for all-cause and CV mortality in general population (Stamler et al., 1999), but evidence regarding the effects of ageing on total and LDL cholesterol levels, and their role as risk factors for CVD and mortality is controversial (Bertolotti et al., 2024; DiNicolantonio & McCarty, 2018). Their levels may rise in older middle age and younger old age but decline with very old age, as well as with chronic disease, inflammation, malnutrition, or poor health status in older persons. In older population, especially in the oldest old, low total cholesterol, and possibly also low LDL cholesterol, may be associated with increased mortality (Nanna et al., 2019; Ravnskov et al., 2016; Schupf et al., 2005; Tikhonoff et al., 2005). An increased cholesterol level still seems to indicate an increased risk for ischemic heart disease mortality also in older populations (McMahon et al., 2007). In another study on our population, the highest age quartile had the lowest total and LDL cholesterol levels, presumably due to general frailty and decreased protein synthesis, and both total and LDL cholesterol had statistically significant predictive ability in primary prevention of atherosclerotic CVD only in men and not in the oldest quartile (Katajamäki et al., 2025). LDL cholesterol may lose its predictive ability in primary prevention in older age, but a meta-analysis showed that lowering LDL cholesterol with medications reduced the risk for major CV events in over 75-year-olds at least as well as in younger adults, especially in secondary prevention (Gencer et al., 2020). Studies have found statin therapy beneficial in older population in secondary prevention of atherosclerotic CVD at least until the age of 80 years and safe as long as recommendations for this population are followed, but side effects may occur more frequently than in the younger population (Félix-Redondo et al., 2013; Bertolotti et al., 2024). Blood pressure is another risk factor whose role may differ in older population compared to younger adult population. Blood pressure rises with age mainly due to increased arterial stiffness. Almost one third of adults worldwide are estimated to have hypertension, and even over 60% of older population (Mills et al., 2016). Hypertension is an independent, strong and direct risk factor for CVD in adult population, and in large epidemiological studies it remains associated with vascular and overall mortality also in old age. High blood pressure is a complex and heterogeneous condition, and older population is also very heterogeneous, which emphasizes the need for individual consideration of the benefits of the treatment of high blood pressure (Lewington et al., 2002). Elisa Heikkilä 20 2.4 Reference intervals 2.4.1 Defining reference intervals Clinicians determine whether an individual's laboratory measurement falls within a normal range based on reference intervals. A reference interval provides a range of acceptable values for a certain analyte (Clinical and Laboratory Standards Institute, 2010). A reference interval is formed by taking an adequate sample from the chosen reference population and defining the intervals so that 95 percent of the values fall within the reference interval. The reference intervals vary in different populations depending on many factors such as age and ethnicity. Also habits of life in the population such as diet and exercise may impact the values in a population. The reference intervals are applied also on people with comorbidities even though reference intervals are usually defined in a healthy population. Also, the methods used in a certain laboratory produce variation in the values. If a laboratory does not define its own reference interval for a certain laboratory analyte but uses the reference interval defined in another population, they should at least verify the reference interval by collecting a sample of reference individuals from their population (Clinical and Laboratory Standards Institute, 2010; Horn & Pesce, 2003). 2.4.2 Older people’s reference intervals Many laboratory values differ in older population from the values on general adult population. Applying reference ranges defined for general adult population may cause overdiagnosis for older population and can lead to treating normal conditions as pathological. The problem has been recognized in various laboratory parameters, and for some such as N-terminal b-type natriuretic propeptide (proBNP) and thyroid hormones there are separate reference ranges according to age. (Fradley et al., 2011; Goudot et al., 2021; Jansen et al., 2024; Nadrowski et al., 2013; Ni et al., 2022.) The reasons for the change seen in various biomarkers with age are not always clear but physiological changes in different organ systems likely change concentrations of some analytes. This natural adaptation should not be considered as pathological (Bech et al., 2017; Duntas, 2021; Gauthier et al., 2020; Mammen, 2019). It is not always clear whether the change seen in laboratory values is due to the normal ageing process or age-related increase in the prevalence of pathologies. Evaluating the distributions of the values may aid in determining whether there is a subpopulation with differing values, which could implicate a need for a separate reference interval for a certain age group. The older the population is, the less there are individuals with no diseases which complicates defining the reference intervals Review of the Literature 21 for older people, and it may be unrealistic to find a healthy reference population in older age groups (Rifai et al. 2018). 2.5 Cardiac markers 2.5.1 Cardiac troponins Cardiac troponins regulate the contraction of striated muscle in the heart. Cardiac troponin T (cTnT) along with cardiac troponin I are expressed specifically by the heart, and they are released from the myocardium when myocytes are damaged which may occur due to a variety of reasons including inflammation, necrosis, or trauma (Askin et al., 2020; Cardinaels et al., 2015; de Lemos et al., 2010; Franzini et al., 2015; Reiter et al., 2011; Wu et al., 2018; Xu et al., 2013). Troponin concentrations may be measured in the peripheral blood in ischemic conditions such as myocardial infarctions and coronary artery disease (de Lemos et al., 2010; Reiter et al., 2011) but also in non-ischemic acute and chronic heart diseases such as left ventricular hypertrophy and heart failure (Aimo et al., 2018; de Lemos et al., 2010; Latini et al., 2007; Wu et al., 2018). Also, chronic kidney disease may increase their concentrations (Askin et al., 2020; Cardinaels et al., 2015; de Lemos et al., 2010; Franzini et al., 2015; Wu et al., 2018; Xu et al., 2013). Cardiac troponins are mainly used for diagnosing or ruling out an acute myocardial infarction or injury (AMI). The diagnosis of an infarction requires the detection of an elevated cardiac troponin value above the 99th percentile upper reference limit. The infarction is considered acute when there is a rise or fall of troponin values indicating a newly detected dynamic state whereas in chronic infarction the troponin concentrations are persistently elevated (Thygesen et al., 2018). Cardiac troponins T and I are the preferred biomarkers for the evaluation of myocardial injury (Thygesen et al., 2007, 2018). Other biomarkers for myocardial injury such as creatine kinase MB isoform are not as sensitive and less specific for myocardial muscle (Goodman et al., 2006). While in younger adults troponin concentrations cannot be measured in a normal state, older people may have detectable concentrations of troponins in peripheral blood without an acute ischemic event. A number of studies show that the level of cTnT increases with age (Cardinaels et al., 2015; de Lemos et al., 2010; Franzini et al., 2015; Li et al., 2015; Menacer et al., 2013; Olivieri et al., 2012; Reiter et al., 2011; Wu et al., 2018; Zhang et al., 2020). The cut-offs used in clinical practice are not based on older population that is highly prevalent in emergency care and other health care appointments. Many studies have found that elevated levels of cTnT are associated with CV and all-cause mortality in general and older populations with or without cardiac Elisa Heikkilä 22 disease (de Lemos et al., 2010; Devereaux et al., 2012; Latini et al., 2007; Neumann et al., 2024, Wu et al., 2018; Xu et al., 2013). Elevation of cTnT was significantly associated with increased all-cause and CV mortality in the US National Health and Nutrition Examination Survey, independently of traditional risk factors, both in non- diabetic and prediabetic individuals. The risk of CVD was even higher in those individuals with both elevated cTnT and prediabetes (Liu et al., 2023). Also, perioperative measurement of cardiac troponins could be used to evaluate a person’s preoperative and postoperative risk when undergoing a noncardiac surgery (Gualandro et al., 2018; Wolfgang, 2013). 2.5.1.1 Troponin assays High-sensitivity troponin assays are recommended and increasingly used (Mair et al., 2018; Thygesen et al., 2018). The high performance of sensitive assays has improved the early diagnosis of myocardial infarctions as they enable cardiac troponin measurement with a high degree of analytical sensitivity and a low level of analytical imprecision at the lower measuring range (Reichlin et al., 2009; Reiter et al., 2011; Thygesen et al., 2012). The increase in the early diagnostic sensitivity of high-sensitivity cardiac troponin assays reduces their specificity for AMI. With the low analytical detection limit of high-sensitivity assays, concentrations of cardiac troponins can be measured in a significant proportion of healthy adults as well, and in most older individuals (Reichlin et al., 2009). Also, more troponin elevations not caused by acute ischemia are detected than with previous less sensitive troponin assays (Mair et al., 2018). 2.5.2 N-terminal b-type natriuretic propeptide Natriuretic peptides are secreted from cardiomyocytes in response to increased wall tension reflecting increased intravascular volume. They promote vasodilatation and regulate salt and water handling (Epstein et al., 1998; Martinez-Rumayor et al., 2008). Natriuretic peptides including proBNP are used for diagnosing and monitoring heart failure (Ponikowski et al., 2016; Taylor et al., 2017). Their concentrations rise in left ventricular dysfunction, but elevated concentrations may also be due to other conditions such as renal impairment or atrial fibrillation (Raymond et al., 2003; Taylor et al., 2017). A background of CVD and especially previous myocardial heart infarction makes heart failure more likely. Natriuretic peptides have important functions in the kidney as they increase renal blood flow and glomerular filtration rate (Volpe, 2014). The increased circulatory blood volume in patients with a chronic kidney disease causes elevation of blood pressure which together with arterial stiffness may lead to cardiac hypertrophy and heart Review of the Literature 23 failure and thus contribute to the elevation in proBNP. The elevation of proBNP in patients with renal failure is partly due to the increased cardiac secretion of proBNP, and partly due to the impaired clearance from the kidneys (Okamoto et al., 2019). Studies have shown that the concentration of proBNP is a predictor of CV events and mortality in general population (Wang et al., 2004), and of future risks in patients with an existing condition such as the risk of mortality after an AMI (Wang et al., 2023), mortality in patients with acute heart failure (Wang et al., 2020) and even stroke risk after a transient ischemic attack (Rodríguez-Castro et al., 2020). There is evidence of the prognostic value of proBNP for CV events on patients in long-term geriatric care (Sheikh Rezaei et al., 2019) and in general older population (Kistorp et al., 2005). The prognostic value of proBNP for mortality after AMI remains even in the oldest patients (Lorgis et al., 2009). The Atherosclerosis Risk in Communities Study found that proBNP could predict all-cause mortality as a single biomarker equally well in non-diabetic older adults as the Pooled Cohorts Equation that includes clinical variables and HDL and total cholesterol, and even better than the equation in diabetic older adults (Wijkman et al., 2022). The concentration of proBNP increases with age (Costello-Boerrigter et al., 2006; Fradley et al., 2011; Galasko et al., 2005; Goudot et al., 2021; Lorgis et al., 2009; Nadrowski et al., 2013). Specific cut-offs to rule-in or rule-out acute heart failure in different age groups have been developed and are widely applied (Januzzi et al., 2006). The selected threshold value has a significant impact on the proportion of results that are considered elevated. 2.6 Frailty Frailty is a syndrome defined as a loss of resources in several domains leading to increased vulnerability to stressors (Clegg et al., 2013; Fried et al., 2001; Pialoux et al., 2012; Rockwood & Mitnitski, 2007). It is considered a multidimensional and complex syndrome including not only physical but also cognitive, social, and psychological domains. When defining the state of frailty of a person, different aspects of health must be considered together. Fried described frailty as a phenotype consisting of five dimensions including unintentional weight loss, weakness, exhaustion, slow gait and low physical activity (Fried et al., 2001). Frailty can also be understood as accumulation of deficits (Rockwood & Mitnitski, 2007). The recovery time increases as people age which increases their vulnerability to stressors, leading to a higher prevalence of frailty (Mitnitski & Rockwood, 2016). There is individual variability in deficit accumulation that modifies the risk for adverse events and death for people of the same age (Rutenberg et al., 2018). At the individual level, the rate of deficit accumulation depends on changes in environmental stresses and variations in the time it takes to recover from stressors (Mitnitski & Rockwood, Elisa Heikkilä 24 2016). The state of frailty is dynamic and potentially reversible (He et al., 2025). Some risk and protective factors for frailty have been identified, and the factors may be biological, environmental, or related to lifestyle (Cohen et al., 2023). Frailty predicts adverse outcomes such as increased falls, hospitalization, morbidity, dependence, and mortality (Clegg et al., 2013; Fried et al., 2001; Hajek et al., 2018). A link has been established between CV health and frailty. Endothelial dysfunction and atherosclerotic lesions may deteriorate circulation and promote physical and cognitive frailty. The relationship between frailty and CVD seems to be bidirectional, and the role of frailty is important when treating older CV patients (Strandberg & Nieminen, 2020). Frailty and even pre-frailty are independent risk factors for CVD (Veronese et al., 2017). 2.7 Frailty indexes Frailty can be measured using FIs that are tools for assessing the individual condition and risk for adverse health events such as mortality, institutionalization, worsening health status, hospitalization, increased falls, morbidity, and dependence of an older person. Symptoms, signs, diseases, disabilities, medications, or laboratory measurements can be combined in an index to measure frailty. (Blodgett et al., 2017; de Vries et al., 2011; Rockwood & Mitnitski, 2007; Song et al., 2010.) For a deficit to be included in a FI, its prevalence must increase with age but not saturate too early. As a group, the deficits must cover a range of systems (Searle et al., 2008). The quality and number of deficits vary in both clinical and laboratory- based indexes. Not all deficits are equal considering the risk they cause but even minor individual deficits that might not singly be significantly related to adverse outcomes together contribute to the person’s state of frailty. The index score is calculated as the proportion of an individual’s deficits in relation to the total amount of deficits chosen for the index (Blodgett et al., 2019; Clegg et al., 2015; Mitnitski et al., 2002; Morley et al., 2013; Rockwood et al., 2004; Rockwood & Mitnitski, 2007; Searle et al., 2008). A review on multidimensional FIs found 611 unique deficits that comprised the FIs with a median of 36 deficits, their number ranging from five to 72. Typically, at least 30 deficits were included. Diseases, most commonly diabetes, hypertension and heart failure, and symptoms or signs such as depression, falls and weight loss, were most used as deficits, followed by disabilities such as hearing or vision loss. Eight percent of the deficits included in the studies in this meta-analysis were abnormal laboratory values (Dlima et al., 2025). Despite the varying deficits that different FIs include, they show a consistent, sub-maximal limit of about 2/3 of the deficits that a person may have (Blodgett et al., 2017; Mitnitski et al., 2015; Ritt et al., 2017; Rockwood & Mitnitski, 2007). Review of the Literature 25 FIs are mostly used for older population, although some have been shown to have predictive value also in younger population. On the other hand, at a very high age, age alone may exceed the predictive ability of a FI (Blodgett et al., 2017). Frailty can be measured by clinical and laboratory-based indexes. The information they give is partly different. Diagnostic methods including laboratory tests are used for detecting or monitoring organ damage more precisely or before it advances and becomes clinically evident. The same applies for FIs, where the clinical parameters give information on the current health status of a person and the person’s functional abilities, and abnormal results in blood tests may reflect preclinical conditions or diseases. Typical characteristics for clinical and self-report FIs, and laboratory-based FIs are non-linearly increasing scores with age, submaximal limits and their associations with mortality. The pattern of deficit accumulation seems to be convex indicating increasing accumulation of deficits with age (Kulminski et al., 2006; Mitnitski et al., 2013, 2015; Rockwood et al., 2007). Clinical FIs have mostly found that women are more frail than men, but they seem to be more resilient to frailty and live longer (Gordon & Hubbard, 2020). The same does not seem to apply for laboratory-based FIs where there is no clear indication of a gender difference in the index scores. Some studies have found men to have higher scores than women, some the opposite or no difference between genders (Hakeem et al., 2023; Sapp et al., 2023). The reason for the paradox of women being more frail but more resilient to frailty is unclear. It has been theorised that women may be more prone to report and seek treatment for health issues than men but likely there are also biological factors involved (Gordon 2019). 2.7.1 Clinical indexes Traditionally FIs are formed using clinical deficits. Some commonly used indexes include FRAIL scale, Rockwood’s FI and PRISMA-7 (de Vries et al., 2011; Hoffmann et al., 2020; Rockwood & Mitnitski, 2007; Song et al., 2010; Thompson et al., 2020). FIs consisting of clinical variables are strongly associated with the risk of death, institutionalization, and worsening health status (Rockwood & Mitnitski, 2007; Searle et al., 2008; Salminen et al., 2019; Viljanen et al., 2021). In earlier studies on our population, clinical frailty tools have been found to be applicable screening instruments among community-dwelling older people where frailty was associated with higher mortality and institutionalization rates according to three different clinical frailty screening tools. Simple frailty tools that comprised of five to seven self-reported items were found comparable with a multidimensional and time-consuming Rockwood’s FI that includes at least thirty items. The hazard ratios (HRs) for different clinical FIs for 10-year-mortality for a person defined as frail were 7.96 for Frail Scale, 5.96 for Rockwood’s FI and 4.41 for PRISMA-7 in unadjusted models, and the Elisa Heikkilä 26 associations persisted after adjustments with age and gender (Salminen et al., 2019). The HRs for institutionalization in unadjusted analyses were 3.32, 8.82, and 3.95, respectively, and the associations remained statistically significant after adjustments (Viljanen et al., 2021). 2.7.2 Laboratory indexes The idea of laboratory-based indexes is that subclinical deficits, taken together, and even including deficits not individually related to mortality, may contribute to the risk of adverse outcomes of ageing. These deficits may precede clinically evident health deficits and have predictive value even in those with little other evidence of frailty. (Blodgett et al., 2016, 2019; Collerton et al., 2012; Howlett et al., 2014; Howlett & Rockwood, 2013; Mitnitski et al., 2015; Rockwood et al., 2015.) The ideal frailty screening tool would quantify frailty based on data that are collected routinely, and that requires minimal participation by patients. Routinely collected laboratory data would fulfill this criterion. In the first study on a laboratory- based FI by Howlett et al., data from the Canadian Study of Health and Aging was re-analyzed leading to the finding that an index based on laboratory deficits could predict mortality. In their study the HRs for mortality increased by 2.8% for each increment in their 23-item laboratory FI. (Howlett et al., 2014.) Research on laboratory-based FIs has increased during the last decade (Hakeem et al., 2023; Sapp et al., 2023). Many studies have demonstrated that prediction of mortality and other adverse health outcomes can be based on laboratory data (Howlett et al., 2014; Rockwood et al., 2015; Blodgett et al., 2016, 2017, 2019; Ritt et al., 2017; Hao et al., 2019; Wang et al., 2022; Ellis et al., 2025). The most studied outcome in laboratory-based index studies has been mortality. In most studies, the laboratory-based indexes were associated also with other outcomes apart from mortality, including disabilities, health care use and self-reported health (Sapp et al., 2023). The index by Howlett et al. that has been used in many studies consisted of 21 laboratory variables and systolic and diastolic blood pressure. The laboratory parameters were albumin, aspartate aminotransferase, calcium, creatinine, folate, folate in erythrocytes, fasting glucose, hemoglobin, mean corpuscular volume, alkaline phosphatase, inorganic phosphorus, potassium, sodium, thyroid stimulating hormone, thyroxine, free thyroxine, urea, vitamin B12, and white blood cell count (Howlett et al., 2014). The parameters and their number are similar in most other laboratory-based indexes. Some have formed indexes consisting of biomarkers that need specialized analytics such as Mitnitski et al. in Newcastle study in over 85- year-olds, where their index of 40 parameters included inflammatory, hematological, Review of the Literature 27 immunological, cell senescence, genetic, and epigenetic markers such as telomere length, DNA repair, and DNA repair to damage ratio (Mitnitski et al., 2015). Rockwood et al. found a strong linear relationship with a laboratory-based index and a clinical FI in a study on older adults in long-term care facilities (Pearson correlation coefficient 0.95). A laboratory-based index could identify long-term care residents at increased risk of death, and it also added to clinical information, as an index combining clinical and laboratory data showed the best predictive value (Rockwood et al., 2015), HRs for the clinical FI 1.03, the laboratory-based FI 1.02, and the combined FI 1.04 for each 0.01 increment in the corresponding FI in age- and gender-adjusted models. Blodgett et al. examined associations of a laboratory- based index and adverse health outcomes in adult population and found that higher index scores were associated with poor health outcomes at all ages (Blodgett et al., 2019). In their study on older men, the correlation between a laboratory-based index and a clinical FI was relatively weak (Pearson correlation coefficient 0.33), but HRs were similar to Rockwood et al.’s study with a slightly higher HR for the clinical FI (1.05) than the laboratory-based FI (1.04), and highest for the combined FI (1.07) for each 0.01 increment in the index score. In their study, rates of institutionalization, doctor visits, medication use, self-reported health and falls increased as each index score increased (Blodgett et al., 2016). The number of laboratory parameters was the same (23) in both studies, but the number of clinical parameters (58 and 39) was different. Blodgett et al. suggested that a laboratory-based index could be utilized as an early screening tool to identify deficit accumulation at the cellular and molecular level before they become clinically visible (Blodgett et al., 2016, 2019). A laboratory-based index has also been studied in acutely ill older adults admitted to hospital, and could be useful also in an acute setting in identifying the severity of illness of patients who are frail at presentation to hospital and for creating their early care plan (Ellis et al., 2020; Jäger et al., 2019; Klausen et al., 2017). A laboratory- based index has also been found predictive of mortality in the very old of over 90 years of age (Hao et al., 2019). 2.7.3 Combined indexes There have been prior studies that have used both clinical and laboratory data to construct FIs (Blodgett et al., 2016, 2017, 2019; Howlett et al., 2014; Ritt et al., 2017; Rockwood et al., 2015). Combining laboratory parameters with self-reported items seems to increase their predictive value, and in most studies where a clinical FI, a laboratory-based FI, and a combination of these indexes was formed, the combined index showed the best predictive ability (Blodgett et al., 2017, 2019; Rockwood et al., 2015; Howlett et al. 2014). Still, a review on laboratory-based FIs found that combining laboratory-based FIs with non-laboratory-based FIs only marginally Elisa Heikkilä 28 increased the predictive ability in seven studies (Hakeem et al., 2023). It is unclear whether the nature of the included deficits or their number increases the predictive ability of the index, since some studies have found that adding more deficits to an FI can strengthen its predictive ability (Blodgett et al., 2019; Rutenberg et al., 2018). 2.8 Institutionalization Institutionalization can be defined differently depending on the country, but in Finland it is considered to include long-term care in an institution, where a person is provided 24/7 assisted living. The rate of institutionalization does not depend solely on the health status of older people themselves but is also affected by health policies. In Finland, the goal is to decrease institutional care by increasing home care to enable older people to live longer at their homes. The care needs for most older people can be met at home but because of the rapidly ageing population, it is still likely that the need for institutional care will increase as well. In addition to home and institutional care, the aim is to increase communal living, which is a lighter form of care for those with impaired function, but who are not in need of 24/7 care. Communal care includes living in a safe and suitable environment with social interactions and services available during daytime, and an individual care plan which may include home care services according to the person’s needs (Ministry of Social Affairs and Health, 2024). Due to the mentioned policies during the last decade, the number of older people in institutional care has decreased in Finland each year, with a twenty percent decrease from 2023 to 2024, being 47 583 persons at the end of year 2024. Four percent of 65-year-olds and older, seven percent of 75-year-olds and older, and sixteen percent of 85-year-olds and older were in institutional care in 2024. The number of people living in assisted or communal care has increased but is still small in comparison to institutional care with 4 529 persons at the end of year 2024 (Finnish Institute for Health and Welfare, 2025). Progressive chronic diseases cause the most need for services including institutional care. Dementia is the leading cause of disability in old age, and it is by far the disease that most often leads to a person’s need to enter long-term care. In over 90-year-old people with dementia, the odds of disability in activities of daily living were almost five-fold compared to people with three other diseases but no dementia (Vargese et al. 2023). The number of disability-adjusted life years from Alzheimer’s disease more than doubled between 2000 and 2021 (World Health Organization, 2024), which contributes significantly to the need for institutional and other care plans. Other factors leading to institutional care besides dementia include multimorbidity, functional dependence, cerebrovascular disease, hip fracture and Parkinson’s disease (Agüero-Torres et al. 2001; Halonen et al. 2019). Review of the Literature 29 2.9 Mortality Noncommunicable diseases have become more prominent causes of death worldwide. CVD is the biggest group of noncommunicable diseases, and caused 19.2 million deaths in 2021, which is 32 percent of all global deaths, followed by malignant neoplasms that caused 9.8 million deaths. Over three quarters of CVD deaths take place in low- and middle-income countries. The number of deaths caused by ischemic heart disease has increased the most, and also the number of deaths caused by Alzheimer’s disease and diabetes has increased. Respiratory infections caused most deaths by a communicable disease with 11.2 million deaths in the world in 2021, out of which COVID-19 caused 8.7 million deaths (World Health Organization, 2024). Also in Finland CVDs as a group cause the most deaths followed by cancers and dementias in general population as well as in older population of 65-year-olds and older. CVD mortality has decreased but it is still the underlying cause of death in about one third of all deaths in Finland. Coronary artery disease causes about half of CVD deaths with similar rates for both men and women. Dementia including Alzheimer’s disease has risen to be the most common single cause of death in Finland causing one fifth of all deaths (Official Statistics Finland, 2024). 30 3 Aims This thesis aimed to create indexes based on routine laboratory analytes and to investigate their associations with mortality and institutionalization, and to study the associations of cTnT and proBNP on all-cause and CV mortality in older population. In addition, we used our study population to create reference limits for cTnT and proBNP in older population. The specific aims were: 1. To analyze whether a laboratory index based on 14 commonly used laboratory tests can be used to evaluate the risk of institutionalization and mortality among Finnish older people during 10- and 18-year follow-ups (Study I) 2. To create indexes to predict the risk for mortality and institutionalization with as few parameters as possible without compromising their predictive ability (Study II) 3. To define reference limits for cTnT and proBNP that would better reflect their concentrations in older people and could be applied in clinical use (Study III) 4. To study the associations of cTnT and proBNP with the incidence of AMIs in an older population with and without previous heart diseases, and their association with CV and all-cause mortality in older population in non-acute conditions (Studies III and IV) 31 4 Material and Methods 4.1 Study population This study is part of the Lieto elderly study, which is a longitudinal epidemiological study carried out in the municipality of Lieto in south-western Finland (Löppönen et al., 2003). All persons born in or prior to the year 1933 (n=1596) were invited to participate in the baseline examination which was carried out between March 1998 and September 1999. Of those eligible, 63 died before they were examined, and 273 refused or did not respond, leaving 1260 (82%) participants, 533 men and 727 women. They were followed for institutionalization and mortality for 18 years for study I, and ten years for study II. For study IV, the participants were followed for ten years for CV and all-cause mortality. Participants no longer living in Lieto at the end of 2016 (n=86) were excluded from the analyses predicting institutionalization, as it was not possible to ascertain whether they were institutionalized in their current municipality or whether they lived at home. Sixty-eight participants were already living in institutional care at the start of the study and were excluded from the institutionalization analyses. After these exclusions, there were 1106 participants in the institutionalization analyses. For study I, participants with missing data of analytes needed for the laboratory index (n=107) were excluded leaving 1153 participants for the final study cohort predicting mortality. Data was missing for 87 of those that were left in the institutionalization analyses, leaving 1019 participants. For study II, participants with missing data of more than five percent of the parameters included in the indexes (n=88 for mortality and n=52 for institutionalization) were excluded leaving 1172 and 1054 participants for the analyses predicting mortality and institutionalization, respectively. For the reference population in study III, all individuals with a kidney disease (International Statistical Classification of Diseases and Related Health Problems 10th Revision [ICD-10] codes N00–N29) at baseline (n=52) or with missing data (n=10) were excluded. Of those without kidney diseases, 435 already had a diagnosis of a heart condition (ICD-10 codes I20–I25, I30–I52 including ischemic heart diseases and other heart diseases) at baseline and were also excluded from the Elisa Heikkilä 32 reference population. This left us with 763 individuals who formed the reference population for cTnT and proBNP. The 435 participants with a previous diagnosis of a heart disease at baseline were used to test the association of the newly created reference ranges with the incidence of AMIs in older population with a previous heart condition. For study IV on all-cause and CV mortality, all individuals with a kidney disease at baseline (n=52) or with missing data (n=10) were excluded from the study population, leaving 1198 participants for mortality analyses. 4.2 Methods 4.2.1 Baseline examination Data for the clinical parameters were gathered from a doctor’s clinical examination including a comprehensive interview and a survey of patient records at baseline. Cross-sectional data collected between March 1998, and September 1999 were used as baseline information. All study participants were clinically carefully examined, including a comprehensive interview with history, lifestyle, and previous diagnoses, Rose questionnaire, numerous laboratory analyses and an electrocardiogram examination (Löppönen et al., 2003). 4.2.2 Laboratory measurements Venous blood samples were obtained with minimal stasis between 8 and 10 am after overnight fast at Lieto Health Centre. All participants were given verbal and written instructions on preparing for the blood sample collection before the laboratory visit. Blood samples were collected, centrifuged and aliquots of serum were stored at -70°C. The analytes that were used for the index studies were analysed from fresh samples at the Central laboratory of Turku University hospital at the time of the baseline examination. The analyses of high sensitivity cTnT and proBNP were performed from previously unthawed stored samples at the Central laboratory of Turku University Hospital during December 2020. The determination of cTnT and proBNP were performed on a Cobas® 8000 e801 analyser using electrochemiluminescence immunoassay (ECLIA) method (Roche Diagnostics, Mannheim, Germany; for cTnT the limit of detection (LoD) 3 ng/L, the limit of quantification (LoQ) (10% coefficient of variation) 13 ng/L, and for proBNP LoD 5 ng/L, LoQ (20% coefficient of variation) 50 ng/L). The mean coefficients of variation for two-level controls were 3.0% and 3.8% for cTnT, and 2.7% and 2.8% for proBNP. Material and Methods 33 4.2.3 Institutionalization Institutionalization was defined as permanent entry into long-term care of which the data were gathered from the municipality’s electronic patient record system and coded by month and year of entry. Eighty-six participants were excluded because they had moved to another municipality during the follow-up period, and their institutionalization could not be ascertained. Sixty-eight participants were already living in institutional care at baseline and excluded from the institutionalization analyses. 4.2.4 Mortality Data from all participants who died by the end of 2016 for study I, and by the end of 2008 for studies II, III and IV were obtained from the Statistics of Finland Causes of Death -registry identified with unique personal identification numbers. In this study, the cause of death was defined as cardiovascular if the death was registered by ICD- 10 codes I10‒I15, I20‒I25 or I30‒I52. 4.2.5 The incidence of acute myocardial infarctions The data on the participants’ diagnoses for AMIs were collected from the municipality’s electronic patient record system, the Finnish Hospital Discharge Register provided by the National Institute of Health and Welfare, and the Finnish Cause of Death Registry, from the Official Statistics Finland, from baseline examination up till the end of 2008, providing a follow-up period of 10 years. The diagnosis for AMI was based on ICD-10 codes I21 and I22. 4.2.6 The laboratory index In study I we created a laboratory index (LI) comprising fourteen laboratory analytes. The laboratory analytes that constitute the LI and their reference ranges or cut-off values are shown in Table 1. The index is calculated as the proportion of individual’s laboratory test results outside reference ranges in relation to the total number of analytes tested. For the index we chose some of the laboratory parameters that were defined at the beginning of the Lieto elderly study. They were selected so that they do not overlap significantly as risk indicators but reflect the health status of different organ systems and that they are routinely tested when examining older patients in Finland. The LI was constructed by coding each analyte as either 0 or 1; 0 indicates that the value was within the normal range or cutoff and 1 that the value was above or below the normal range. The sum of these values was then divided by the total number of the analytes resulting in a score ranging from 0 to 1 for each individual. Elisa Heikkilä 34 To compare the adverse outcomes of individuals with different LI scores, we divided the participants in five categories (1. LI ≤0.08 [≤1 laboratory test result outside reference ranges], 2. LI 0.09–0.14 [2 laboratory test results outside reference ranges], 3. LI 0.15–0.21 [3 laboratory test results outside reference ranges], 4. LI 0.22–0.42 [4 to 5 laboratory test results outside reference ranges], and 5. LI ≥0.43 [≥6 laboratory test results outside reference ranges]). Table 1. Reference ranges used for the analytes in the laboratory index for men and women Men Women Hemoglobin (g/L) 128-168 117-153 Albumin (g/L) 36.1-47.5 34.8-46.1 Calcium (mmol/L) 2.17-2.47 2.17-2.47 Urate (µmol/L) 180-420 130-340 TSH (mU/L) 0.4-4.5 0.4-4.5 Creatinine (µmol/L) < 135 < 125 Ferritin (µg/L) 20-240 10-100 CRP (mg/L) < 3 < 3 Sodium (mmol/l) 136-144 136-144 Potassium (mmol/L) 3.5-4.8 3.5-4.8 Glucose (mmol/L) 4.0-6.4 4.0-6.4 ALT (U/L) < 50 < 35 ALP (U/L) < 300 < 300 LDL cholesterol (mmol/L) < 3.5 < 3.5 Abbreviations: TSH, thyroid stimulating hormone; CRP, C-reactive protein; ALT, alanine amino- transferase; ALP, alkaline phosphatase; LDL, low-density lipoprotein 4.2.7 The combined clinical and laboratory index For study II, we combined the 36 clinical and 14 laboratory parameters that have been used in our earlier studies separately, to form an index with possibly a better predicting ability on mortality and institutionalization. The clinical parameters included 36 symptoms, signs, and disabilities similarly to other FIs (Mitnitski et al., 2001; Searle et al., 2008), and are shown in Table 2. The laboratory parameters were the same as in study I. The combined index (LI+FI) was constructed similarly to the LI by coding each deficit or laboratory analyte as either 0 or 1 and dividing the sum of these values by the total number of the parameters resulting in a score ranging from 0 to 1. For this study, we used established cut-points that have been used in FIs (Song et al., 2010) (1. LI+FI ≤0.085, 2. LI+FI 0.085–0.2499, 3. LI+FI ≥0.25) and thus divided the participants in these three categories. Material and Methods 35 Table 2. The clinical parameters used in the combined indexes Needs help with toileting 1 2 Needs help with dressing and undressing 1 2 Needs help with preparing meals 1 2 Needs help with house work 1 2 Needs help with heavy household chores 1 2 Needs help with personal care 1 2 Needs help with moving about inside house 1 2 Arthritis or rheumatism High blood pressure 1 Chronic bronchitis or emphysema Diabetes mellitus 1 Heart diseasea 1 Cancer 1 Stomach or intestinal ulcers Suffers from the effect of stroke 1 Urinary incontinence 2 Stool incontinence 2 Shortness of breath 1 2 Angina pectoris 1 Other medical problemsb 1 No regular physical exercise 1 2 Vision problem 1 2 Hearing problem 1 2 Feeling hopeless 1 Emotional problem 2 Memory problem 1 2 Bodily pain 2 Speech problem 1 Resting tremor 1 Five or more medications 1 2 Difficulties carrying or lifting light loads 1 2 Mobility problem 1 2 Limited kind of amount of activity 1 2 Feeling tired all the time 1 2 Weight loss 2 aKnown heart disease at baseline (ICD-10: I20-I25, I48, I49) bOther disease at baseline (ICD-10: E03, E05, G20, G35, J44-J46, M15-M17, M47) 1 Independently associated with 10-year-mortality in univariable analysis 2 Independently associated with 10-year-institutionalization in univariable analysis Elisa Heikkilä 36 4.3 Statistical analyses HRs and their 95% confidence intervals for institutionalization in studies I and II, all-cause mortality in studies I, II and IV, CV mortality in study IV and the incidence of AMIs in study III were calculated using Cox proportional hazard models. Proportional hazards assumption was tested using Martingale residuals. The follow- up periods were calculated from the baseline examination and measurements to the end of the follow-up period of 10 or 18 years or to the death of the individual. Death was used as a competitive factor in the Cox regression analyses for institutionalization and in the analyses for AMIs. For studies I and II, unadjusted and age- and gender-adjusted analyses were conducted. For study IV for the analyses of the associations of cTnT and proBNP with all-cause and CV mortality, the analyses were adjusted also for other statistically significant baseline characteristics. P-values less than 0.05 were considered statistically significant. All statistical analyses were performed using SAS System for Windows, version 9.4 (SAS Institute Inc., Cary, NC, USA). The formation of the reduced indexes in study II was started by selecting those clinical and laboratory parameters that were associated with institutionalization or mortality in the univariable Cox regression analysis. There were no significant correlations between different laboratory parameters but some of the clinical parameters correlated significantly and were left out from the reduced indexes. A backward stepwise Cox regression analysis (exclusion criteria, p≥0.05) was performed to identify the parameters which best predicted mortality and institutionalization and to reduce the number of parameters in the index (Collett, 2014). The reduced indexes were also calculated by including age (64‒69, 70‒74, 75‒79, 80 or over) and gender to the indexes. Clinical and laboratory parameters were scored with 0 or 1 points, female gender was scored with 1 point for the institutionalization index and male gender with 1 point for the mortality index, and age was scored with 0 points for ages 64 to 69, 1 point for ages 70 to 74, 2 points for ages 75 to 79 and 3 points for the age of 80 or over, based on the parameter estimates of the stepwise Cox regression model and statistical testing in the effect of different weights on the index. This resulted in a range of 0 to 7 for the combined index to predict institutionalization and a range of 0 to 13 to predict mortality. Receiver- operating characteristic curves were used to define the cut-off points for the reduced indexes to predict institutionalization and mortality (Gonen, 2007). The optimal cut- off points were chosen using Youden index (Youden, 1950). Sensitivity, specificity, positive predictive value and negative predictive value for cut-off points were calculated. Kaplan-Meier survival curves for institutionalization and mortality during the 10-year follow-up were implemented to compare participants below and above the cut-off points (Collett, 2014). Material and Methods 37 For the creation of reference limits for cTnT and proBNP, Mann-Whitney U-test was used to compare the concentrations between age groups and genders. The distributions of cTnT and proBNP values were skewed, and therefore log- transformed for reference limit calculations. Normal distribution method was used to calculate 97.5% reference limits (mean+1.96× standard deviation [SD]) with their corresponding 95% confidence intervals after exclusion of outliers according to the 3 SD criterion. For cTnT we defined also the 99% reference limit (mean+2.33×SD) that is recommended for the diagnosis of AMI (Thygesen et al., 2018). After the reference limits were calculated, the values were transformed back to the original units. In study IV, the mortality of the highest and lowest quartiles of the analytes were compared to study the associations of cTnT and proBNP with all-cause and CV mortality in each group. 4.4 Ethics The Lieto Elderly Study was conducted according to the guidelines of the Declaration of Helsinki. The Ethics Committee of the Hospital District of Southwest Finland approved the study protocol (Diary number 112/1802/2015). Participants provided written informed consent for the study. 38 5 Results 5.1 Characteristics of the participants The mean age of the participants (n=1259) at baseline (1998‒1999) was 73.5 years (SD 6.8, range 64‒100), 532 were men (42 %), and 727 women (58 %). Most study participants (88%) had basic or less than basic level of education defined as six years of elementary school, and only 12% more than basic education. Most participants (72%) had a Mini Mental State Examination score of more than 25 points. Most participants were living with someone, and 30% were living alone. Some participants were excluded from each study resulting in slightly different numbers of participants (Figure 4). Baseline characteristics for the participants for study I are shown in Table 3. Results 39 Figure 4. Flow chart of the study participants. Abbreviations: LI, laboratory index; FI+LI, combined clinical and laboratory index; CV, cardiovascular. n=1596 invited 1260 participants 63 died before examination 273 refused / did not respond 68 already in institutional care 86 no longer living in Lieto at the end of the follow-up period 107 with missing data for the LI 88 with missing data for the FI+LI Study I Institutionalization analyses n=1019 Mortality analyses n=1153 Study II Institutionalization analyses n=1054 Mortality analyses n=1172 52 with a kidney disease at baseline 10 with missing data for cardiac markers 435 of those with a diagnosis of a heart condition at baseline Study III Reference population n=763 Study IV CV and all-cause mortality analyses n=1198 Elisa Heikkilä 40 Table 3. Baseline characteristics of study participants for mortality analyses in study I (n=1153) n (%) Age, mean (SD), range 73.6 (6.8), 64‒100 Age 64‒74 720 (62) 75‒84 334 (29) ≥85 99 (9) Women 663 (58) Living alone 344 (30) Education Basica or less than basic 1050 (91) More than basic 103 (9) Mini Mental State Examination ≤26 323 (28) Body mass index, kg/m2 <20 68 (6) 20‒24.9 317 (28) 25‒29.9 498 (43) 30‒34.9 206 (18) ≥35 61 (5) Number of prescribed medicines <5 855 (74) 5‒7 207 (18) 8‒9 66 (6) ≥10 25 (2) aSix years of elementary school 5.2 Index studies 5.2.1 The laboratory index 5.2.1.1 Prediction of mortality (Study I) Altogether, 422 (37%) and 806 (70%) subjects deceased during the 10- and 18-year follow-ups, respectively. Higher LI predicted increased mortality. Index scores of 0.09 or over and 0.15 or over, predicted increased mortality during the 10- and 18-year follow-ups, respectively (Tables 4 and 5). These associations persisted after adjustments for age and gender. Figure 5 shows Kaplan-Meier survival curves by the categories of LI. Results 41 Table 4. Hazard ratios (HR) and their 95% confidence intervals (CI) (in parentheses) of the laboratory index (LI) for mortality during the 10-year follow-up Total n Deceased n (%) Unadjusted HR (95% CI) P-value Adjusteda HR (95% CI) P-value LI ≤0.08b 383 92 (24) 1 0.09‒0.14c 326 119 (36) 1.61 (1.23‒2.12) <0.001 1.69 (1.28‒2.22) <0.001 0.15‒0.21d 211 87 (41) 1.96 (1.46‒2.63) <0.001 1.84 (1.37‒2.48) <0.001 0.22‒0.42e 190 93 (49) 2.46 (1.84‒3.28) <0.001 2.20 (1.64‒2.94) <0.001 ≥0.43f 43 31 (72) 5.56 (3.71‒8.40) <0.001 3.75 (2.46‒5.72) <0.001 Total 1153 422 (37) aValues are adjusted for age and gender The number of laboratory values outside reference ranges by categories of LI: b0–0.08 ≤1 c0.09–0.14 =2 d0.15–0.21 =3 e0.22–0.42 =4–5 f0.43– ≥6 Table 5. Hazard ratios (HR) and their 95% confidence intervals (CI) (in parentheses) of the laboratory index (LI) for mortality during the 18-year follow-up Total n Deceased n (%) Unadjusted HR (95% CI) P-value Adjusteda HR (95% CI) P-value LI ≤0.08b 383 242 (63) 1 0.09‒0.14c 326 213 (65) 1.14 (0.95‒1.37) 0.169 1.17 (0.98‒1.41) 0.091 0.15‒0.21d 211 160 (76) 1.51 (1.23‒1.84) <0.001 1.47 (1.20‒1.80) <0.001 0.22‒0.42e 190 152 (80) 1.80 (1.47‒2.20) <0.001 1.72 (1.40‒2.12) <0.001 ≥0.43f 43 39 (81) 3.41 (2.43‒4.79) <0.001 2.93 (2.07‒4.16) <0.001 Total 1153 806 (70) aValues are adjusted for age and gender The number of laboratory values outside reference ranges by categories of LI: b0–0.08 ≤1 c0.09–0.14 =2 d0.15–0.21 =3 e0.22–0.42 =4–5 f0.43– ≥6 Elisa Heikkilä 42 Figure 5. Survival curves by the laboratory index (LI) comprising 14 commonly used laboratory tests during the 18-year follow-up. Reprinted from Study I (Heikkilä et al., 2021). The number of laboratory values outside reference ranges by categories of LI: 0–0.08 ≤1 0.09–0.14 =2 0.15–0.21 =3 0.22–0.42 =4–5 0.43– ≥6 5.2.1.2 Prediction of institutionalization (Study I) Altogether, 151 (15%) and 305 (30%) subjects were institutionalized during the 10- and 18-year follow-ups, respectively. The LI did not significantly predict institutionalization during either of the follow-ups in unadjusted or age- and gender- adjusted models (data not shown). 5.2.2 The combined index 5.2.2.1 Prediction of institutionalization (Study II) Altogether 149 (14%) subjects were institutionalized during the 10-year follow-up. Higher index score of the 50-parameter combined index was associated with an increased risk of institutionalization in the 10-year follow-up. Both groups with LI+FI 0.085‒0.2499 and LI+FI 0.25 or over had statistically significantly higher risk of institutionalization compared to the group with LI+FI less than 0.085 with HRs 2.42 and 6.64, respectively. These associations persisted after adjustments for age and gender. Results 43 5.2.2.2 Prediction of mortality (Study II) Altogether 413 (35%) subjects deceased during the 10-year follow-up. Higher index score of the 50-parameter combined index was associated with an increased risk of death in the 10-year follow-up. Both groups with LI+FI 0.085‒ 0.2499 and LI+FI 0.25 or over had a statistically significantly higher risk of death compared to the group with LI+FI less than 0.085, with HRs 2.23 and 9.39, respectively. These associations remained statistically significant when adjusted for age and gender. 5.2.3 The reduced indexes 5.2.3.1 Prediction of institutionalization (Study II) Twenty-two clinical parameters from the large, combined index were independently associated with 10-year-institutionalization (Table 2). Hemoglobin was the only laboratory analyte that was statistically significantly associated with institutionalization. After a backward variable selection in the Cox regression model, the parameters that were left for the reduced index to predict institutionalization were need for help with preparing meals, need for help with heavy household chores and need for help with moving about inside house. Female gender and increasing age were significant predicting factors for institutionalization and adding these factors to the index improved its predictive ability (Table 6). Each factor was coded either 0 or 1; 0 indicates that the person did not need help with the task and 1 that the person needed help. One point was given for female gender and 0 to 3 points for increasing age resulting in a range of 0 to 7 for the index score. The parameters and their scoring can be seen in Table 7. The best cut-off limit for the increased risk of institutionalization was found to be ≥4 points. The reduced index with the chosen optimal cut-off limit showed a much greater sensitivity in comparison to the 50-parameter combined index with the cut-off limit 0.25 habitually used for FIs. Using this cut-off limit, the specificity of the larger index was higher than the specificity of the reduced index. The largest area under the curve (0.78) was found for the reduced index that included age and gender with a sensitivity of 74.7% and a specificity of 69.4% for the institutionalization index (Table 6). The higher the index score, the higher percentage of our study population were institutionalized during the 10-year-follow-up period. This can be seen in figure 6a. Figure 7a shows Kaplan-Meier survival curve by the cut-off limit of 4 points. Elisa Heikkilä 44 Table 6. AUC-values, cut-off limits and their sensitivities, specificities and positive and negative predictive values for the 50-parameter combined index and for the reduced indexes in predicting 10-year institutionalization and mortality AUC (95 % CI) Cut-off limit (scale) Sensitivity (%) Specificity (%) PPV (%) NPV (%) Institutionalization 50-parameter combined index 0.71 (0.66‒0.75) ≥0.25* (0‒1) 38.3 84.9 29.4 89.3 Reduced index without age and gender 0.74 (0.70‒0.78) ≥2** (0‒3) 71.0 71.8 30.2 93.5 Reduced index with age and gender 0.78 (0.75‒0.82) ≥4** (0‒7) 74.7 69.4 29.6 94.1 Mortality 50-parameter combined index 0.77 (0.74‒0.80) ≥0.25* (0‒1) 42.5 92.0 74.2 74.7 Reduced index without age and gender 0.79 (0.77‒0.82) ≥5** (0‒9) 61.8 84.2 69.5 79.2 Reduced index with age and gender 0.83 (0.80‒0.85) ≥8** (0‒13) 77.0 75.8 64.9 85.1 *Habitual cut-off limit for a frailty index **Best cut-off limit defined by Youden index Abbreviations: AUC, area under the curve; CI, confidence interval; PPV, positive predictive value; NPV, negative predictive value 5.2.3.2 Prediction of mortality (Study II) Twenty-seven clinical parameters and eleven laboratory parameters from the large, combined index were independently associated with 10-year-mortality (Table 2). The laboratory analytes associated with mortality were hemoglobin, albumin, calcium, urate, TSH, alanine aminotransferase, creatinine, ferritin, CRP, sodium, potassium, glucose and alkaline phosphatase. The reduced index for mortality included three laboratory analytes and six clinical parameters. The parameters were elevated or decreased blood hemoglobin value, elevated plasma CRP level, and elevated or decreased plasma sodium level, the need for help with preparing meals, the need for help with heavy household chores, difficulties carrying or lifting light loads, limited kind of amount of activity, diabetes mellitus and heart disease. One point was given for each of these deficits or a laboratory value outside the reference ranges. Increasing age and male gender were significant predicting factors for mortality so, similarly to the index predicting institutionalization, 0 to 3 points was given for age and 1 point for male gender. This resulted in a score ranging from 0 to 13 with the best cut-off limit being ≥8 points for the prediction of a person being at an increased risk of death. The parameters and their scoring can be seen in Table 7. Results 45 Increasing index score increased deaths in the study population during the 10-year- follow-up period as seen in figure 6b. Figure 7b shows the Kaplan-Meier survival curve by the cut-off limit of 8 points. When comparing the reduced and large combined indexes, sensitivity was higher for the reduced index using the chosen cut-off limit, but specificity higher for the larger index. The reduced index that included age and gender showed the largest area under the curve (0.83) with a sensitivity of 77.0% and a specificity of 75.8% for the mortality index (Table 6). Table 7. The parameters and their scoring for the reduced indexes Index points Institutionalization Needs help with preparing meals 1 Needs help with heavy household chores 1 Needs help with moving about inside house 1 Female gender 1 Age 70 to 74 1 Age 75 to 79 2 Age 80 or more 3 Total index score 0–7 Mortality Elevated or decreased blood hemoglobin value 1 Elevated plasma c-reactive protein level 1 Elevated or decreased plasma sodium level 1 Needs help with preparing meals 1 Needs help with heavy household chores 1 Difficulties carrying or lifting light loads 1 Limited kind of amount of activity 1 Diabetes mellitus 1 Heart disease 1 Male gender 1 Age 70 to 74 1 Age 75 to 79 2 Age 80 or more 3 Total index score 0–13 Elisa Heikkilä 46 Figure 6. Institutionalization (A) and mortality (B) rates by each index score during the 10-year follow-up. Reprinted from Study II (Heikkilä et al., 2023). A B n=164 n=303 n=228 n=157 n=113 n=88 n=33 n=18 0 10 20 30 40 50 60 70 80 90 100 0 1 2 3 4 5 6 7 Institutionalization (%) n=63 n=202 n=205 n=186 n=135 n=104 n=79 n=72 n=55 n=36 n=19 n=5 n=2 0 10 20 30 40 50 60 70 80 90 100 0 1 2 3 4 5 6 7 8 9 10 11 12 Mortality (%) Results 47 Figure 7. Rates of institutionalization by the reduced index with the cut-off score ≥4 (A) and mortality by the reduced index with the cut-off score ≥8 (B) during the 10-year follow-up. Age and gender are included in both reduced indexes. Reprinted from Study II (Heikkilä et al., 2023). Elisa Heikkilä 48 5.3 Cardiac marker studies 5.3.1 Baseline characteristics of the reference population (Study III) The reference population was formed by 763 individuals, 308 men and 455 women aged over 64 years. The mean age of the men was 72.1 years (SD 6.2), and the mean age of the women 72.8 years (SD 6.4). 435 participants with a previous heart disease were excluded from the reference population, 193 men, whose mean age was 73.5 years (SD 6.4) and 242 women, whose mean age was 76.3 years (SD 7.5). The median concentrations of cTnT among the reference population were 10.9 ng/L (interquartile range [IQR] 8.2‒15.2) for men, and 9.2 ng/L (IQR 6.5‒12.2) for women. The median concentrations of cTnT among the population with a previous heart disease were 14.1 ng/L (IQR 10.1‒22.1), and 10.7 ng/L (IQR 7.8‒19.0), respectively. The median concentrations of proBNP among the reference population were 81 ng/L (IQR 47‒153) for men and 123 ng/L (IQR 75‒210) for women. The median concentrations of proBNP among the population with a previous heart disease were 201 ng/L (IQR 94‒617), and 259 ng/L (IQR 116‒635), respectively. The median concentrations of cTnT and proBNP in different age groups among the reference population are shown in Table 8. The relation of cTnT and proBNP concentrations with age in the reference population can be seen in Figure 8. Table 8. Median concentrations and interquartile ranges (IQR) of cardiac troponin T (cTnT) and N-terminal natriuretic b-type propeptide (proBNP) in different age groups among the reference population n Median concentration of cTnT (ng/L) IQR Median concentration of proBNP (ng/L) IQR Men 64 to 69 years 138 9.0 6.9‒12.1 64 34‒118 70 to 74 years 85 10.7 8.3‒13.4 75 47‒121 74 to 79 years 49 14.6 10.7‒19.4 115 79‒186 Over 80 years 38 16.7 13.3‒31.2 131 66‒239 Women 64 to 69 years 176 6.9 5.3‒9.5 91 59‒158 70 to 74 years 130 8.8 6.8‒11.2 113 76‒170 74 to 79 years 77 10.4 7.7‒13.9 133 103‒214 Over 80 years 76 15.3 10.9‒21.7 219 125‒355 Results 49 Figure 8. The relation of cardiac troponin T (cTnT) in men (A) and women (B), and N-terminal natriuretic b-type propeptide (proBNP) in men (C) and women (D) with age in the reference population. Reprinted from Study III (Heikkilä et al., 2024). 5.3.2 Reference limits for troponin T and N-terminal b-type natriuretic propeptide (Study III) The reference limits were defined separately for both genders in four age groups because of statistically significant age group and gender differences in cTnT and proBNP values. For cTnT, we calculated a 99% reference limit of 25 ng/L for men aged 64 to 69 years, 28 ng/l for men aged 70 to 74 years, 38 for men aged 75 to 79 years, and 71 for men aged 80 years and older, and 18 ng/L, 22 ng/l, 26 ng/l, and 52 ng/L for women, respectively. The 97.5% reference limits for proBNP were 272 ng/L, 287 ng/l, 373 ng/l and 686 ng/L for men, and 341 ng/L, 377 ng/l, 471 ng/l, and 794 ng/L for women, respectively (Table 9). Ta bl e 9. R ef er en ce li m its a nd th ei r 9 5% c on fid en ce in te rv al s (C I) fo r c ar di ac tr op on in T (c Tn T) a nd N -te rm in al n at riu re tic b -ty pe p ro pe pt id e (p ro BN P) cT nT pr oB N P n 97 .5 % re fe re nc e lim it 95 % C I f or re fe re nc e lim it 99 % re fe re nc e lim it 95 % C I f or re fe re nc e lim it n 97 .5 % re fe re nc e lim it 95 % C I f or re fe re nc e lim it M en 64 to 6 9 ye ar s 13 6 22 19 ‒2 4 25 22 ‒2 9 13 3 27 2 21 7‒ 34 2 70 to 7 4 ye ar s 82 24 21 ‒2 8 28 24 ‒3 2 78 28 7 21 7‒ 38 0 74 to 7 9 ye ar s 48 32 27 ‒3 9 38 31 ‒4 6 44 37 3 27 3‒ 50 9 80 y ea rs a nd o ld er 37 58 42 ‒7 9 71 52 ‒9 7 31 68 6 41 7‒ 11 30 W om en 64 to 6 9 ye ar s 17 4 16 14 ‒1 8 18 17 ‒2 0 16 8 34 1 28 5‒ 40 9 70 to 7 4 ye ar s 12 4 19 17 ‒2 1 22 19 ‒2 4 12 2 37 7 31 2‒ 45 5 74 to 7 9 ye ar s 75 23 19 ‒2 7 26 23 ‒3 1 73 47 1 36 6‒ 60 7 80 y ea rs a nd o ld er 74 43 35 ‒5 3 52 42 ‒6 3 67 79 4 60 3‒ 10 45 Elisa Heikkilä 50 Results 51 5.3.3 The associations of troponin T and N-terminal b-type natriuretic propeptide with the ten-year incidence of acute myocardial infarctions (Study III) There were 109 new AMIs in the whole study population during the 10-year follow- up period, and 53 of those in the population who already had a previous diagnosis of a heart disease. Out of the 401 participants with a heart disease and a normal cTnT level at baseline 46 (11%), and out of the 34 participants with a heart disease and an elevated cTnT seven (21%) had an AMI during the follow-up period. There were 299 participants with a heart disease who had a normal proBNP level at baseline and 29 (10%) of those had an AMI, whereas there were 136 participants with a heart disease and an elevated proBNP level and 24 (18%) of those had an AMI during the follow-up period. According to the Cox regression analyses, there were no statistically significant associations of elevated cTnT above the 99% reference limit with the incidence of AMIs during the 10-year follow-up period among participants with and without a previous heart disease, HRs 2.0 (0.88–4.56) and 2.20 (0.80–6.06), respectively. Elevated proBNP was statistically significantly associated with the incidence of AMIs both among participants with and without a previous heart disease, HRs 1.95 (1.13–3.35) and 2.54 (1.25–5.17), respectively. 5.3.4 The associations of troponin T and N-terminal b-type natriuretic propeptide with cardiac and all-cause mortality (Study IV) Of the 1198 participants in this study, 676 had a CVD, 430 of those had hypertension, and 254 had coronary artery disease. 174 participants had diabetes. The median concentration of cTnT was 11.7 ng/L for men (IQR 8.7‒17.9) and 9.7 ng/L for women (IQR 6.9‒14.6). The median concentration of proBNP was 117 ng/L for men (IQR 63‒240) and 148 ng/L for women (IQR 84‒300). A total of 467 (37 %) subjects deceased during the 10-year follow-up period, and 149 of those of a CVD. There were statistically significant age group and gender differences in both cTnT and proBNP concentrations. We divided the population into 64‒74-year-olds, and over 75-year-olds because cTnT (median 8.9 ng/L vs 15.5 ng/L, p<.001) and proBNP (median 104 ng/L vs 230 ng/L, p<.001) increased significantly with age. Women had lower concentrations of cTnT (p<.001) and higher concentrations of proBNP (p<.001), and genders were analysed separately. Both elevated cTnT and proBNP concentrations were statistically significantly associated with CV and all-cause mortality during the 10-year follow-up period in Elisa Heikkilä 52 each group without adjustments. cTnT was statistically significantly associated with CV and all-cause mortality in over 75-year-old men and women when adjusted for the statistically significant baseline factors, but in the younger age group statistical significance was not reached. ProBNP was significantly associated with CV mortality in both age groups in men and women also when adjusted for the baseline factors, and in over 75-year-old men and 64‒74-year-old women for all-cause mortality. 53 6 Discussion This thesis provides new indexes for clinical use for the prediction of institutionalization and mortality in a population aged 65 years and older. Other main results include the confirmation of the predictive value of cTnT and proBNP for CV events and CV and all-cause mortality in older Finnish population. We defined reference limits on our representative older study population and confirmed the previous finding that the conventional reference limits are not valid for older population. The indexes created in this thesis may aid clinicians in identifying those at high risk of adverse health outcomes. To be of value, indexes must be easy enough to apply in use, which is why we tried to reduce the number of parameters in the indexes. Earlier, there have been FIs mostly comprising clinical deficits that are used for the prediction of adverse outcomes including institutionalization and mortality. Different previously formed clinical FIs have been shown to apply also in our study population (Salminen et al., 2019; Viljanen et al., 2021). Indexes to predict frailty and adverse outcomes have also been based on laboratory parameters instead of clinical findings (Howlett et al., 2014; Blodgett et al., 2019). Some studies have found strong and some weak correlations between clinical FIs and laboratory-based indexes (Blodgett et al., 2016; Howlett et al., 2014; Ritt et al., 2017; Rockwood et al., 2015; Yang et al., 2018). Whereas clinical findings may be easy to obtain for an experienced clinician based on a clinical examination and a thorough interview, limitations in time and differences in both doctor’s and patient’s subjective assessment may complicate this evaluation. A distinct feature of a laboratory-based index compared with a clinical FI is that it may offer an opportunity to find deficits at cellular, molecular, or organ level prior to their clinical manifestations (Blodgett et al., 2017). Clinical and laboratory-based indexes are partly different entities, and do not always define the individual’s risk similarly, although both can be used to predict mortality. On the other hand, also different clinical FIs define partly different people as frail (Salminen et al., 2019; Viljanen et al., 2021). Studies have already proven the predictive ability of laboratory-based FIs, but the number of laboratory parameters in most FIs has been overly large to apply them Elisa Heikkilä 54 in use, for example, on a routine heath care visit of an older person. Especially considering the increasing number of older people, more simple and less time- consuming tools are needed, such as the fourteen-parameter LI that was created as part of this thesis study. The laboratory parameters used in the LI, as in most other formerly studied laboratory-based FIs, are measured from venous blood samples. Taking a venous blood sample can be considered an invasive procedure but with minimal risks of complications. Blood samples are commonly drawn prior or after a doctor’s appointment so defining the analytes needed for the index would rarely require an additional laboratory visit. Also, the point of our index was to choose mostly analytes that are commonly examined as part of a health evaluation. The index score could be calculated automatically by laboratory information systems. cTnT and proBNP are both proteins secreted or released by the myocardium (de Lemos et al., 2010; Martinez-Rumayor et al., 2008), and they can be used as biomarkers. Their elevated concentrations indicate an increased risk of CV mortality, but their predictive value can also be seen in their associations with total mortality in our study, and in earlier studies (de Lemos et al., 2010; Devereaux et al., 2012; Latini et al., 2007; Wang et al., 2004; Wang et al., 2019; Wu et al., 2018; Xu et al., 2013). Their concentrations increase with age, and separate but varying reference limits are in use for older people for proBNP, but not for cTnT. We defined new reference limits for both analytes in our study population. There was a significant rise in both analytes with age. Nevertheless, there were associations of increased concentrations of both markers measured in a non-acute situation and the incidence of future AMIs and with all-cause and CV mortality. 6.1 Study population The data came from a community-based representative sample of Finnish older population, with a high participation rate (82%), and a long follow-up period. The study population consisted of a broad sample of the aged population living in the mid-sized semi-industrialized municipality of Lieto. At the time of the baseline examination, the percentage of the population aged over 65 years was 12 in Lieto, and 15 in the whole country. From the beginning of the study, the age structure in Finland has changed significantly with currently 24 percent of the population aged 65 years or over. The estimation for 2050 is 26 percent, and for 2070 31 percent (Official Statistics Finland, 2025). The gender distribution of the participants is comparable to the distribution of this age group in the whole country (Official Statistics Finland, 2025). Comparable age and gender structure at the time of the baseline examinations enables generalizability of the results to most Finnish municipalities. Nevertheless, Finland is a vast country with differences in population in different parts of the country such as the prevalence of CVDs in western and Discussion 55 eastern parts of the country (Finnish Institute for Health and Welfare, 2022). The reasons for these differences are unclear but likely due to both genetic differences and lifestyle factors such as the prevalence of cigarette smoking (Nuotio et al., 2020). These factors may lead to differences on how the created indexes and biomarkers predict outcomes in a certain population. Also, apart from the change of age structure, many other factors such as life habits and the prevalence of diseases may have changed from the time when this study was started. Testing the results in another population would be needed to ensure generalizability of the results to a specific population. 6.2 Findings 6.2.1 Laboratory index (Study I) The results from study I suggest that an index based on fourteen routine laboratory analytes can be used to predict mortality in an older population. The LI was significantly associated with mortality but not with institutionalization during the 10- and 18-year follow-ups. The association of the LI with mortality remained after adjustments for age and gender. Fourteen analytes is a small number of parameters compared to prior studies on laboratory-based FIs (Blodgett et al., 2017; Howlett et al., 2014; Ritt et al., 2017; Rockwood et al., 2015). In a systematic review and meta-analysis conducted in 2023 containing 38 studies on laboratory-based FIs, the number of the parameters ranged from the 14 parameters in our study to 77 parameters, with the average of 30 parameters. 93 percent of the included parameters were laboratory parameters. They found no difference in HRs between indexes including 20 to 25 parameters to those including 26 to 77 parameters, which supports the idea that less parameters may be enough if they reflect the health status of a variety of physiological systems (Sapp et al., 2023). In another review also conducted in 2023, 51 percent of the laboratory-based FIs included some non-laboratory parameters, most often pulse or blood pressure (Hakeem et al., 2023). In studies with a larger number of analytes, some analytes reflected the health status of the same or partly same organ system such as hemoglobin, red blood cells, mean corpuscular volume and hematocrit (hematopoiesis), or alanine aminotransferase, aspartate alanine transferase and gamma-glutamyl transferase (the liver) (Howlett et al., 2014; Ritt et al., 2017). In selecting the analytes that construct the LI in our study, care was taken that the information obtained from the analytes did not overlap significantly but captured information with respect to health status of different organ systems. Based on our study, it seems that fourteen analytes is Elisa Heikkilä 56 enough to achieve predictive value for mortality when the analytes are selected appropriately. 6.2.1.1 Associations of the laboratory parameters with mortality Studies have found associations with mortality for all the fourteen analytes individually. Some of them have been more thoroughly studied than others, and some of them have strong evidence on their association with mortality. Especially anemia and inflammation are known to be risk factors for increased mortality, and thus a decreased hemoglobin level and an elevated CRP level have been found to be strongly associated with increased mortality. Anemia increases mortality from all causes in the older population, and the lowest mortality has been found at normal hemoglobin levels (Culleton et al., 2006; Shavelle et al., 2012). Anemia is also associated with reductions in functional capacity and quality of life. Anemia becomes progressively more common with age and is prevalent in both men and women over the age of 65 years. Common causes for anemia in older people are nutritional deficiencies and chronic diseases. Anemia is related to many diseases, but it seems to be related to ageing as well as it is estimated that one third of anemia in older patients is unexplained (Guralnik et al., 2022). Anemia related to ageing with no apparent underlying cause is likely multifactorial, with variable contributions from renal dysfunction, endocrine deficiency causing weaker response to erythropoietin, chronic inflammation, androgen deficiency and possibly beginning myelodysplasia (Guralnik et al., 2022). Elevated levels of CRP have been shown to predict an increased risk of all-cause and CV mortality in the general population (Harris et al., 1999; Li et al., 2017). CRP levels are correlated with various CV risk factors, such as age, smoking status, hypertension, body mass index, frequency of exercise, and multiple lipid and non- lipid plasma-based indicators (Rohde et al., 1999). Ageing presents with systemic low grade chronic inflammation, and concentrations of many circulating inflammatory mediators rise with age (De Maeyer & Chambers, 2021; Harris et al., 1999). Chronic inflammation causes dysfunction of the immune system and strongly predicts morbidity and mortality. Increasing age is associated with increased incidence of infections and cancer, and decreased vaccine efficacy. This increased morbidity observed with age is believed to be due in part to a decline in adaptive immunity, termed immunosenescence (De Maeyer & Chambers, 2021). Inflammation is also a strong predictor for frailty, and the frailest have the highest levels of circulating CRP as well as other inflammatory mediators (Giovannini et al., 2011). In addition to hemoglobin and CRP, deviating levels of each of the other analytes of the LI have associations with mortality. Creatinine is commonly used to estimate Discussion 57 glomerular filtration rate of the kidney that is associated with the risk of all-cause mortality (Willey et al., 2020). Kidney function slowly declines with age. The decline gradually impairs glomerular filtration and kidney’s regulatory range of function that affects matching sodium and potassium excretion to dietary intake. Even when the kidney function declines, it is sufficiently preserved in healthy older people under normal conditions, but the homeostasis is challenged in extreme conditions or diseases (Gekle, 2017). Studies have found associations with sodium levels and mortality in general population (Corona et al., 2013; Mohan et al., 2013), hospitalized patients (Waikar et al., 2009), and in patients with a chronic kidney disease (Kovesdy et al., 2012). Even a moderate serum sodium decrease was found to be associated with an increased risk of mortality in commonly observed clinical conditions. Hyponatremia develops in many conditions and is the most common electrolyte disorder in hospitalized patients. It is more common among aged patients, and with patients with hypertension, diabetes, coronary artery disease, stroke, chronic obstructive pulmonary disease, cancer, and psychiatric disorders, but remains associated with the risk of mortality when adjusted for age and comorbidities. A chronic hypotonic state beyond that of the underlying illness seems to affect the mortality risk (Corona et al., 2013). Also hypo- and hyperkalemia are important electrolyte abnormalities that can both cause cardiac arrhythmias, and especially in patients with CV or renal disease. Low or high potassium levels are associated with increased mortality, but the risk is significantly greater in patients with comorbid conditions (Collins et al., 2017). Especially hyperkalemia, but also hypokalemia has been found to be associated with increased mortality especially in patients with hypertension (Krogager et al., 2017), chronic kidney disease, diabetes mellitus, or heart failure, and in older population (Collins et al., 2017). Possibly high and low potassium levels even within reference ranges may increase the risk for mortality, but the exact potassium levels associated with increased risk remain controversial (Collins et al., 2017; Krogager et al., 2017). Increased or decreased concentrations of urate acid are associated with increased CV and all-cause mortality, also in over 65-year-olds. Especially hyperuricemia increases the risk of mortality, but the association is U-shaped (Tseng et al., 2018). Hypocalcemia may be associated with increased mortality, especially in the older population with comorbidities (Xu et al., 2025). An increased mortality rate has been observed in older individuals with thyroid dysfunction. The level of thyroid stimulating hormone increases with healthy ageing, which is important to note in diagnosing thyroid disorders in older people (Mitrou et al., 2011). Elisa Heikkilä 58 There is also some evidence on the association of ferritin, which is both a critical protein in iron homeostasis, low in iron deficiency, and an acute phase inflammatory marker, with mortality (Ellervik et al., 2014; Mahroum et al., 2022). Increased serum alkaline phosphatase levels have been linked with increased all- cause mortality in general population, and alanine aminotransferase levels also in older population (Schmilovitz-Weiss et al., 2018; Yan et al., 2023). The somewhat controversial role of LDL cholesterol in the older population has been discussed in the chapter of CVD risk factors. 6.2.1.2 Laboratory parameters in prediction of institutionalization Clinical FIs have been shown to predict institutionalization (Rockwood & Mitnitski, 2007), also in our study population (Viljanen et al., 2021), but the LI did not. This finding can be considered logical as the most common causes of institutionalization, dementia and cognitive impairment, cannot be predicted by routine laboratory testing (Bharucha et al., 2004; Gnjidic et al., 2012; Luppa et al., 2010; Salminen et al., 2017, 2020; von Bonsdorff et al., 2006). The same applies for some other factors that are known to predict institutionalization, impairing an older person’s ability to live independently; increased falls, decreasing body mass index (Salminen et al., 2017, 2020) and functional impairment and disabilities, especially when combined with cognitive impairment (Gnjidic et al., 2012; Luppa et al., 2010; von Bonsdorff et al., 2006). A systematic review on laboratory-based indexes found an association of a laboratory-based index with institutionalization in one study, but not in another study apart from our study (Hakeem et al., 2023). In the study where there was an association between the laboratory-based index and institutionalization, it was weaker than for the clinical FI. Also, their index included blood pressure and pulse in addition to laboratory parameters (Blodgett et al., 2016). 6.2.2 Combined indexes (Study II) As clinical and laboratory-based indexes measure frailty differently, it could be ideal to combine both to acquire the most accurate predictive ability. We created an index that combines our clinical and laboratory indexes that had both been studied independently. Both had been found to predict mortality in our population cohort, and the clinical FI also institutionalization. The clinical data included deficits that reflect several aspects of the person’s health status such as mobility, daily functions and diagnosed diseases. In addition to the LI+FI with all the fourteen laboratory and fifty clinical parameters, we used a statistical method to find a combination of a few parameters Discussion 59 that best predicted mortality and institutionalization out of the large number of variables. This way we could create indexes that take little time to implement and still have a good predictive ability. 6.2.2.1 Prediction of institutionalization The LI+FI was not better for predicting institutionalization compared to the clinical FI alone that was studied in a previous study on the same population (Viljanen et al., 2021). After implementing the statistical method to reduce the number of parameters to only those that showed the best predictive value, no laboratory parameters were left in the index predicting institutionalization, which is consistent with the result that the LI could not predict institutionalization. The three parameters that remained in the reduced index for institutionalization were need for help with preparing meals, need for help with heavy household chores, and need for help with moving about inside the house. Being able to prepare meals independently is one of the instrumental activities of daily living, that reflect the person’s ability to carry out daily activities, and their cognitive function. Earlier studies have shown dependence in daily activities to be a predictor of morbidity and mortality in older populations (Millán-Calenti et al., 2010). Need for help with moving about inside house is also consistent with the earlier finding that self- reported walking ability predicts a person’s risk for institutionalization (Viljanen et al., 2021). 6.2.2.2 Prediction of mortality The LI+FI was better at predicting mortality than the clinical FI or the LI alone. The clinical parameters left in the reduced index as predicting factors for mortality were need for help with preparing meals, limited kind of amount of activity, need for help with heavy household chores and difficulties carrying or lifting light loads. Two of them reflect the physical capacity of the person. Many studies have found physical activity to have a positive effect on health and to reduce mortality (Hupin et al., 2015; Lear et al., 2017). Limited kind of amount of activity was a predicting factor for mortality and refers also to other than physical activity, including interests and hobbies. In addition to these, diabetes and heart disease were left as predicting factors in the index predicting mortality. The laboratory parameters that were found most significant in predicting mortality and were left in the reduced index were hemoglobin, CRP and sodium, all independently associated with mortality according to previous studies (Corona et al., 2013; Culleton et al., 2006; Hupin et al., 2015; Lear et al., 2017; Li et al., 2017; Millán-Calenti et al., 2010; Shavelle et al., 2012). Elisa Heikkilä 60 6.2.2.3 The predictive value of the combined and the reduced indexes The results showed that a large index with fifty parameters included many unnecessary parameters that did not increase its predictive value and therefore could be replaced with a reduced index with only a few carefully chosen parameters that were individually associated with institutionalization or death. The reduced indexes had even a slightly better predictive ability in comparison with the 50-parameter combined index. These indexes with only a few clinical questions in addition to three basic laboratory tests would take very little time in a doctor’s appointment and thus could be used to screen older people. This kind of short intervention could be done during any health care contact with an older person. The indexes could also potentially be calculated automatically if the necessary information was collected in the electronic patient records similarly to electronic FIs that are automatically populated from routine collected data contained within the electronic patient records (Clegg et al., 2016). 6.2.3 Reference limits for troponin T and N-terminal b-type natriuretic propeptide in older population (Study III) Comorbidities are especially common among older population, which complicates the definition of reference limits in this population. In accordance with previous studies we found that cTnT and proBNP concentrations are higher in older population without any cardiac symptoms or diagnosis of a cardiac disease when compared to general adult population (Cardinaels et al., 2015; Costello-Boerrigter et al., 2006; de Lemos et al., 2010; Fradley et al., 2011; Franzini et al., 2015; Galasko et al., 2005; Goudot et al., 2021; Li et al., 2015; Menacer et al., 2013; Nadrowski et al., 2013; Olivieri et al., 2012; Reiter et al., 2011; Wu et al., 2018; Zhang et al., 2020). In our study population the concentrations of cTnT were higher in older men in comparison with older women in all age groups, as most other studies have found as well (de Lemos et al., 2010; Olivieri et al., 2012; Zhang et al., 2020). Women had higher concentrations of proBNP as in earlier studies (Costello-Boerrigter et al., 2006; Fradley et al., 2011; Galasko et al., 2005). We established reference limits for cTnT and proBNP for older people to help decision-making in clinical settings. We chose those individuals with no diagnosed renal diseases or CVDs for the reference population instead of requiring the individuals to be without any illnesses. This is both much easier to implement and closer to the situation in real life. Discussion 61 6.2.3.1 Cardiac troponin T Especially the median cTnT values in our population with a previous heart disease were close to the cut-off of 14 ng/L recommended by the manufacturer to rule-out AMI, even though none had acute symptoms at the time of the initial examination. The established reference limits were higher than 14 ng/L in each age group and increased with age. Misdiagnosis causes harm both for the individual and the society. Overdiagnosis is an important issue that may nowadays be the most important issue when considering the diagnostics of AMIs because of more sensitive diagnostics of troponin rise. Even if no AMIs should be missed, the costs with overdiagnoses are non-negligible, and the resources could be used for alternative health benefits (McCarthy et al., 2024, Eggers et al. 2009). The improved laboratory diagnostics that enable us to find many more patients with positive test results than previously with the simultaneous increase in life expectancy and the percentage of older population brings new challenges, making it ever more important to apply appropriate reference ranges for persons with different ages. 6.2.3.2 N-terminal b-type natriuretic propeptide The reference limits we established for proBNP in our population increased significantly with age. There are several threshold levels that are used for proBNP, such as the National Institute for Health and Care Excellence guideline of 400 ng/L likely to rule out heart failure, and the European Society of Cardiology guideline that recommends further investigation at proBNP levels above 125 ng/L in the non-acute setting, and above 300 mg/L in acute settings (Ponikowski et al., 2016; National Guideline Centre (UK), 2018). According to the manufacturer of the proBNP assay the levels of proBNP rise with increasing age, and in their reference group from the Gutenberg Health study there was a 97.5th percentile value of 879 ng/L for men aged 65 to 74 years and 623 ng/L for women aged 65 to 74 years in a population with no prevalent CVD (Tzikas et al., 2013). The 97.5th percentile values in our population were lower (272, 287, 373 and 686 ng/L for men depending on the age group, and 341, 377, 471 and 794 ng/L for women). There is a large variation between the thresholds recommended by guidelines. 6.2.3.3 Age groups As cTnT and proBNP levels clearly increase with ageing, we defined separate reference ranges for age groups 64 to 69-year-olds, 70 to 74-year-olds, 75 to 80-year- olds, and for 80-year-olds and older. The concentrations of both cTnT and proBNP seem to rise with advancing age also after the age of 80 but the dispersion in the Elisa Heikkilä 62 population gets larger. This might be due to more underlying asymptomatic heart conditions in the oldest old population even if in our study all participants were carefully examined to exclude all persons with any signs or symptoms of a cardiac disease so that only 60.5 percent of the initial population formed the reference group. The possibility remains that some individuals had asymptomatic cardiac conditions that were not diagnostic at the time of the baseline examination, but these undiagnosed underlying heart conditions were at a stable stage at the time. As this might be a more relevant problem in the eldest group with more dispersion, we suggest using the reference limits of over 80-year-olds for all persons over the age of 80, even if it is likely that the reference limits continue to rise with advancing age. 6.2.4 Prediction of cardiovascular events and cardiovascular and all-cause mortality with troponin T and N-terminal b-type natriuretic propeptide (Studies III and IV) We studied if elevated cTnT of proBNP could predict future AMIs in our study participants who already had a diagnosis of a heart disease at baseline. Participants with a heart disease and an elevated cTnT level were more likely to have a new AMI, and only three of those were both alive and did not have a new AMI by the end of the follow-up period suggesting that a person with a previous heart disease and an elevated cTnT level is at a significant risk of a new AMI and death. It is nevertheless important to note that most of those who had a new AMI, had a level of cTnT at baseline which was inside our newly established reference limits. This finding suggests that while those with an elevated cTnT level are at higher risk, lower cTnT level cannot rule out a future risk for an AMI. Elevated proBNP level was very common in patients with a previous diagnosis of a heart disease. Elevated levels of proBNP have been shown to predict CV events and mortality in older people with or without heart failure and also after AMI (Lorgis et al., 2009; Wang et al., 2004). In our study the participants with an elevated level of proBNP were more likely to have an AMI both among participants with and without a previous diagnosis of a heart disease. The results suggest that the levels of both cTnT and proBNP reflect the health status of the heart even when no acute symptoms are present. The results are in accordance with previous research. Both studied cardiac biomarkers have been found to predict incident coronary heart disease, also in a multiethnic population without known CVD and independent of established risk factors and ethnicity. The Multi-Ethnic Study of Atherosclerosis also found that adding a baseline proBNP improved the risk discrimination of established CV risk scores, the Framingham Risk Score and the Pooled Cohort Risk Equation for estimation of 10-year CVD risk. In this study, proBNP predicted incident CVD also Discussion 63 independently above these risk scores. The use of proBNP on top of clinical risk factors resulted in a net 20% improvement in the classification of CVD risk (Daniels et al., 2015). Risk models based on traditional risk factors provide only moderate discrimination of a person’s risk for CV events, and especially for those with an intermediate risk score (Leong et al., 2025). Identifying and integrating other factors that include circulating biomarkers could improve risk assessment accuracy and strengthen primary prevention efforts. Emerging research indicates that the levels of cTnT, proBNP and also CRP in the general population may indicate subclinical heart damage and adverse cardiometabolic health, providing further insights into an individual's long-term CVD risk (Neumann et al., 2025). These biomarkers have been shown to have incremental value when adding them into models containing established risk factors (Neumann et al., 2024; Westermann et al., 2017; Willeit et al., 2017). While the predictive value of conventional risk factors seems to attenuate in older people, the added value of cTnT or proBNP in risk estimation seems to be largest in older people (Neumann et al., 2024). In our study, the CV and all-cause mortality were higher in participants with concentrations of cTnT or proBNP at the highest quartile compared with those in the lowest quartile, although statistical significance was not reached in all groups. Our results suggest that the concentrations of both cTnT and proBNP reflect the cardiac health status, even in the absence of acute symptoms. Age obviously plays a major role in the mortality of older adults, which is why analyses in age groups and adjustments with age are important. In our study, the associations of higher cTnT and proBNP concentrations with both CV and all-cause mortality remained. The associations were mostly stronger in the older group of over 75-year-olds than in the younger age group. Overall, the associations were stronger for CV mortality. 6.3 Strengths and limitations The strengths of this study include the large sample size, good participation rate of 82%, and a long follow-up period that enable broad generalizability of the results. The data comes from a community-based representative sample of the Finnish population. In the first study on the LI, we used two follow-up periods of 10 and 18 years, which can be considered especially long, as it was the longest follow-up period in a review of 38 studies on laboratory-based indexes (Sapp et al., 2023). In the analyses regarding institutionalization, there was missing data for those who no longer lived in Lieto at the end of the study. Also, persons still living at home after the follow-up period were considered not institutionalized, which can be Elisa Heikkilä 64 considered a limitation since some of them can still be institutionalized during their lifetime. Other than these limitations, the data for the outcomes can be considered reliable. The dates for institutionalization were gathered from electronic patient systems which is more exact compared to earlier studies in which the information was gathered by interview (Hajek et al., 2015; Bravell et al., 2009). The data for mortality is reliable as all deaths in Finland are registered in the Official Finnish Cause of Death Registry. The diagnoses were collected from multiple data sources: the municipality’s electronic patient record system, the Finnish Hospital Discharge Register, and the Finnish Cause of Death Registry. Since the data was cross-sectional, we were unable to investigate if subclinical deficits were detectable prior to clinical ones. The clinical parameters in the combined FI+LI were from a previously validated FI that consists of parameters describing a person’s functional abilities and some chronic diseases. It can be considered a limitation to the study that the index does not contain some other factors that also alter a person’s risk, such as cigarette smoking or body mass index. These factors have typically not been directly included in indexes that assess frailty but can also contribute to a person’s risk for mortality and other adverse outcomes. Nevertheless, these factors may have caused diseases that are included in the index. The laboratory parameters for the LI were chosen based on an educated guess so that they would be informative of the state of various organ systems and easily obtainable in clinical use, but we did not study how the inclusion of other laboratory parameters would have altered the predictive ability of the index. All the parameters were classified as binary variables as a person either having or not having a certain deficit without considering how abnormal the result was. The laboratory parameters were classified as either inside or outside reference ranges and not as continuous variables. A certain value is considered normal of abnormal depending on the chosen reference ranges. Those with a value close to the cut-off limit are classified as having a certain deficit or not, although a single value close to the cut-off limit may be due to normal variation. The inclusion of multiple parameters in the indexes lessens the effect of a single laboratory value, but especially in the reduced indexes, the effect of single parameters classified as normal or abnormal is larger. We did not separate if the value was below or above the normal range. This simplifies the clinical usage of the index as both low and high values are considered abnormal, even if naturally they implicate different disease states. While this approach does not provide information on the pathology causing the laboratory abnormality, it is typical in indexes that predict frailty and other non-disease-specific states. All the created indexes should be validated in another population. Discussion 65 The reference ranges for cTnT and proBNP were defined from a community- based representative sample of older population. The sample size was relatively large, even if the oldest age groups of men and women were smaller which is often a challenge when establishing reference ranges for older population. cTnT and proBNP were measured only at one time-point at baseline. The health status of the participants was based on a clinical examination by a physician, patient health records and a comprehensive interview but cardiac imaging was not performed for the reference population. We formed the reference population with individuals without symptoms and with no known heart conditions. Diagnosed cardiac conditions are common in older population but possibly also undiagnosed underlying cardiac conditions which may cause dispersion in the concentrations of cTnT and proBNP, especially in the oldest. It is reasonable to assume that potential hidden undiagnosed co-morbidities may impact the reference ranges, and coronary artery disease cannot be excluded without angiographic examination. Increased cTnT levels may be detected in conditions other than acute ischemia such as inflammation of the heart, endothelial dysfunction, micro-vascular disease or left ventricular strain (Agewall et al., 2011; Askin et al., 2020). It is possible that also the increase in all-cause mortality that we found in our study with high cTnT and proBNP levels, is partly related to undiagnosed and untreated myocardial damage that might increase the likelihood of non-CV death. Renal function is important in metabolism and removal of both biomarkers (Chaulin, 2022; Martinez-Rumayor et al., 2008), and therefore underlying decreased kidney function may influence the concentrations of cTnT and proBNP, although in this study the individuals with kidney diseases at baseline were excluded. cTnT presents in different forms in plasma due to its degradation. Intact and long forms of cTnT are detected early after AMI, and truncated smaller fragments of cTnT in myocardial injury attributable to other causes. The heavily truncated fragments may be responsible for chronic cTnT elevations as seen for example in renal dysfunction (Cardinaels et al., 2013; Aalto et al., 2026; Airaksinen et al., 2022; Mingels et al., 2017). The commercial cTnT assay that was used to define cTnT levels in our study measures all forms of cTnT so we could not separate if the participants’ cTnT elevations were caused by intact or fragmented forms of cTnT. Separation of the different forms of cTnT could have brought more insight on the causes of the elevations of cTnT levels even though participants with a renal dysfunction were not included in the study. Also, we did not study macro troponins that could have resulted in increased concentrations of cTnT in some individuals. All the laboratory parameters were measured at one time point. A serial measurement might improve the predictive value of the biomarkers, as was found on a study on proBNP, where a significant rise or drop of at least 25 percent in three years was associated with an increase or a decrease in the incidence of CVD, Elisa Heikkilä 66 respectively (Daniels et al., 2015). A similar result was found earlier in older adults in the Cardiovascular Health Study (deFilippi et al., 2010). 6.4 Implications for clinics and future research 6.4.1 Implications for clinics In clinical settings, the implementation of an index consisting of available routine laboratory data may be easier and more harmonized than using data based on clinical assessment. Some of the risk factors that can be identified by laboratory tests can be treated when found early enough. Others reflect a person’s health status but cannot be treated as such. For example, a decreased or an elevated sodium level is not a disease-specific finding, but as part of an index can be indicative of an increased risk. An elevated CRP level indicates silent inflammation and the risk for atherothrombosis and ischemic heart disease (The Emerging Risk Factors Collaboration, 2012; Giovannini et al., 2011; Kistorp et al., 2005). Elevated levels of cTnT and proBNP indicate an increased risk for CVD and mortality, and even if ways to lower their levels are limited, their elevated levels indicate a need to affect other treatable risk factors for CVD. Using the LI would not be time-consuming or expensive; it could serve as an alert for the clinician to pay attention to those patients with a high index score. A laboratory-based index is an objective approach for assessing frailty with easy interpretation. Apart from the smaller than normal number of deficits, the LI that was created as part of this thesis fulfills the criteria that is generally considered when creating a FI: the variables are diverse, representing different aspects of health, they are scored 0 or 1, their prevalence increases with age, reflecting deterioration of health, but they do not saturate so that every person at old age would have a certain deficit. On the other hand, a combination of clinical and laboratory parameters might be an optimal solution for the most comprehensive assessment of an older individual’s frailty status. An index with a small number of parameters could be practical to use as part of an older person’s any health care visit; in a health check-up, emergency care, or when planning follow-up care after a hospital stay or a surgery. In large indexes, some parameters, such as the different activities of daily living, may correlate significantly with each other, and thus, their number can be reduced without compromising the predictive ability of the index. Identification of increased risk may aid in assessing the health status of an older person, and planning their health interventions, although the right interventions must be considered on an individual basis. A person’s cTnT and proBNP concentrations must be interpreted in connection with the person’s age, overall health status, and Discussion 67 existing health conditions. Interventions directed at treating those with elevated cTnT or proBNP levels may reduce morbidity and mortality from CV and other causes. The important question is how to treat a patient after defining their index score. The aged population is very diverse, and individuals with the same chronological age might have a very different biological age. The indexes could aid in defining the individual’s condition that is not based on their chronological age. This could help in targeting interventions, building an individual care plan, and considering which treatments are appropriate for the individual. The parameters of the clinical FIs are mostly descriptive of a person’s functional status. When a person already has disabilities, primary prevention is late but further decline can still be affected and there is evidence that an FI score can also improve (He et al., 2025; Mitnitski et al., 2012; Ng et al. 2015). The idea behind laboratory-based indexes is to detect also subclinical defects before their clinical manifestations and thus they could possibly be used also in prevention of clinical defects. Sometimes underlying diseases causing laboratory defects may be found. The index score helps to identify those at high risk, and then to target interventions to delay adverse outcomes, even if the individual parameters cannot always be treated. Assessing the prognosis of geriatric patients is essential for clinicians to balance the potential risks and benefits of available treatment options. The index score could aid in determining the possible benefit that an individual may gain from medications that reduce the risk for certain illnesses such as lipid-lowering or anti-thrombotic medications. It may be uncertain if a very frail person benefits from a medical intervention, such as a medication that reduces a risk that might actualize in a few years. A very high index score could aid in determining if aggressive examinations or treatments are sensible when an individual’s life expectancy is short. There are still no ready answers to questions on what index score a person must have to be suitable for a certain intervention or treatment. Lifestyle intervention may be considered at any age. There is evidence on the benefits of lifestyle changes such as smoking cessation or increasing physical exercise also in old age. Probably the biggest issue is how to motivate a person to change their habits. The finding that the FI scores can also improve indicate the abilities of organisms to repair (He et al., 2025; Mitnitski et al., 2012). Physical, nutritional, and cognitive interventions can aid in reversing frailty (Ng et al., 2015). Frailty is closely connected to CVD. Screening for frailty status is important also to prevent CVD as CV events contribute significantly to disability and mortality in older people. Frailty increases vulnerability to adverse CV events. On the other hand, the CV system affects the health of all systems of the organism and its defects predispose the person to frailty. Also, many of the risk factors for these conditions are the same. Interventions that could be used for treating frailty, such as increasing Elisa Heikkilä 68 physical activity and nutritional interventions, are also useful for preventing the onset of CVD (Strandberg & Nieminen, 2020; Veronese et al., 2017). The ESC guideline from 2021 recognizes a gap of knowledge on management of CVD risk in older people over 85 years of age, although cessation of smoking and increasing physical activity are considered important (Visseren et al., 2021). CVD is also a major risk factor for the development of dementia, and treating the modifiable risk factors for CVD also contributes to the prevention of dementia (Kauko et al., 2024). Assessment of a person’s risk for CVD is essential to reduce CV events, dementia, institutionalization and mortality, and to increase functional life years. In Finland, both CVD prevalence and mortality have decreased because of improved life habits and better prevention and treatment of risk factors, and treatment of CVD (Jousilahti et al., 2016). Nevertheless, the increasing older population and improved treatment may increase the global prevalence of coronary artery disease and other CVDs in the future. This study confirmed that cTnT and proBNP can be used in screening for long- term CVD and mortality risk of older people and identifying a subgroup of older people that need health care interventions. Baseline concentrations of cTnT and proBNP were higher in older population than in younger population without CVDs, which complicates the diagnostics in acute situations. Using age-adjusted reference limits may be useful to improve the diagnostics of CVDs. The application of separate reference limits for older population would have clinical implications, considering the high prevalence of older patients in emergency care as well as in other health care appointments. Troponin levels are often examined with admittance of an older person in an emergency room, and it is important to notice that it is quite likely that they may have their cTnT levels higher than the conventional cut-off limit of 14 ng/L also without an acute ischemic disease. When the reference limits that are established for general adult population are used for older people, their higher cTnT and proBNP concentrations may cause overdiagnosis and lead to unnecessary examinations and treatments. In older patients with elevated levels of cTnT, the change in cTnT concentrations between two following samples is of clinical significance in acute situations, but less follow-up samples and unnecessary monitoring at emergency care are needed when the initial cut-off limits are more appropriate for the population. Even if follow-up samples are still needed when an older person presents with acute cardiac symptoms, it is important to understand that their baseline cTnT level may be over the conventional cut-off limit unrelated to the acute symptoms. 6.4.2 Implications for future research By using a statistical method to reduce the number of parameters of the combined clinical and laboratory index, starting from fifty, they were reduced to nine for the Discussion 69 mortality prediction index, and three for the institutionalization prediction index. This very significant reduction of parameters could be done without compromising the predictive ability of the combined indexes. The parameters for the LI were selected so that they would mostly reflect the health status of different organ systems. Nevertheless, it is possible that all of the fourteen parameters are not necessary, and their number could still be reduced. There is some evidence on the predictive value of each of the fourteen analytes for mortality, at least in some settings, but it may be that all these parameters are not necessary to include in the index. Also, our LI included LDL cholesterol whose predictive role in the mortality of older population is not clear. For example, it would be interesting to see how well a combination of only hemoglobin, sodium and CRP, that were the only laboratory analytes left for the reduced combination index, predicts mortality. In this thesis, mortality and institutionalization were used as outcomes in the index studies, and CV events, CV mortality, and total mortality in the cardiac marker studies. As the main goal is to increase healthy and functional life years, it would be useful and interesting to test how these newly-created indexes predict other outcomes, and to find early indicators for functional decline in future research. The new reference ranges for cTnT and proBNP should be tested in another population of people aged 65 years and older. 70 7 Conclusions This thesis shows that older population free of heart and kidney diseases have higher levels of cTnT and proBNP compared to general adult population and thus we suggest using separate reference limits for older population. The new suggested reference limits divided in four age groups for men and women should be validated in another population of older people. Using these new cut-off limits, elevated proBNP was associated with the occurrence of AMIs in older population with and without a previous diagnosis of a heart disease during a 10-year follow-up period. Older people with higher cTnT and proBNP concentrations have an increased risk of CV and all-cause mortality, even when they do not indicate an acute CV event. Acknowledging the elevated risk may aid in targeting follow-up, prevention, and treatment adequately and more individually. Findings of this thesis show that a practical index based on 14 routine laboratory tests can be used to predict mortality among older people. The number of routine laboratory test results outside reference ranges correlates with older people’s mortality. A laboratory-based FI could be an easily applicable screening instrument in clinical settings. Laboratory and clinical deficits can also be combined to form an index to predict an individual’s risk for mortality, which may be the optimal solution because clinical and laboratory deficits provide different information. Nevertheless, laboratory parameters cannot aid in prediction of institutionalization. According to this thesis a large number of parameters, as typically included in clinical and laboratory-based FIs, is not necessary, and their number could be significantly reduced by statistical methods without compromising the predictive ability of the indexes. 71 Acknowledgements This study was conducted at the department of Clinical Chemistry at the University of Turku during the years 2018 to 2026. I want to thank my supervisors whose knowledge on their fields I admire. Adjunct Professor Marika Salminen has facilitated this project and guided me through challenging phases of my research project. Her expertise combined with her good spirit always left me more encouraged after our meetings, and I am truly thankful for having her as my supervisor. Adjunct Professor Laura Viikari took time for our meetings alongside her busy clinical work as a geriatrician, and with her clinical expertise always wanted to make sure we are concentrating on relevant issues to best serve clinicians and patients. I thank Professor Kari Pulkki for his valuable feedback on my articles. I am indebted to Adjunct Professor Kerttu Irjala, who has been an inspiration for me and other junior researchers. She is the driving force in this whole project and her passion for research as well as her genuine will to encourage and aid us forward is impressive. I am thankful also for my other co-writers, whose comments have been thoughtful and improved my work. Taina Katajamäki and Anna Viljanen have proceeded their work alongside mine. I want to show my gratitude to statistician Tero Vahlberg and data manager Teemu Kemppainen for their aid in the statistics for this project. I wish to thank also all of those who have worked on the Lieto study before I started my project. Professor Tiina Laatikainen and Associate Professor Hanna-Maria Roitto are acknowledged for the time and effort they gave to reviewing this thesis. Much has happened in my life during this lengthy period of working on this thesis. I have had the pleasure to get to know and work with many skilful people while specializing in Clinical Chemistry. My time specializing would not have been the same if I could not have shared it with Nikolai Huttunen and Terhi Pallonen, along with all other doctors and chemists, with whom I have shared countless inspiring conversations, consultations and cups of coffee. All of our team have helped me both grow and find joy in my work. I wish to thank all my senior colleagues both in the fields of hematology and chemistry, who have teached me along the way, and the heads of the department. I am happy to share my working Elisa Heikkilä 72 days with Joanna Danielsson, Saija Sirén, and all my current co-workers at Special Chemistry, and with all my friends at Tyks Laboratories. I want to express my deepest thanks for my mother, Anne and my father, Pekka. I am grateful for seeing you as loving grandparents for my children to whom you give priority in your lives. I am happy to have grown with my dearest siblings, Aleksi and Sonja. I can count on you, and I wish to always be there for you. I am grateful for having my mother-in-law, Eeva, in our family, and for the care you have given to our children. I am blessed with friends that I can share my life with. My cousin Linda has shared countless experiences with me all my life. My oldest friends from pre-school and school times remain close to me. The times we get together mean the world to me. My friends from medical school I hold in great value in my life. My every day is enlightened by my boys. It has been overwhelming to see Otava grow to a thoughtful and empathetic little person who is each day finding more courage to take new steps in his own life even if I still get to see you wave to me though the pre-school window each morning. Onni lives every day to the fullest with often overwhelming emotions of a four-year-old, and with great enthusiasm and concentration in what you engage in. Little Edvin is the joy to us all. Even if, being only one year old, sometimes we must restrict the implementation of your creative ideas, your eagerness to build puzzles and legos, try on and walk in everybody’s shoes, and participate in your brothers’ artwork, is admirable. Finally, I wish to thank my closest person, my husband Jussi. You have lived difficult times with me, and shared the most precious moments life has given us. Turku, February 2026 Elisa Heikkilä 73 References Aalto R., Lahtinen A., Airaksinen K., Vasankari T., Hellman T., Koskimäki L., Wittfooth S. (2026). Design and Analytical Evaluation of an Immunoassay for Long Forms of Cardiac Troponin T. J Appl Lab Med., 11(1):36-47. doi: 10.1093/jalm/jfaf143 Abdelhafiz, A. H., & Sinclair, A. J. (2022). Diabetes in the elderly. Medicine, 50(11), 737–740. https://doi.org/10.1016/j.mpmed.2022.08.008 Abdellatif, M., Rainer, P. P., Sedej, S., & Kroemer, G. (2023). Hallmarks of cardiovascular ageing. Nature Reviews Cardiology, 20(11), 754–777. https://doi.org/10.1038/S41569-023-00881-3 Abrignani, M. G., Lucà, F., Abrignani, V., Pelaggi, G., Aiello, A., Colivicchi, F., Fattirolli, F., Gulizia, M. M., Nardi, F., Pino, P. G., Parrini, I., & Rao, C. M. (2024). A Look at Primary and Secondary Prevention in the Elderly: The Two Sides of the Same Coin. Journal of Clinical Medicine, 13(15), 4350. https://doi.org/10.3390/jcm13154350 Agewall, S., Giannitsis, E., Jernberg, T., & Katus, H. (2011). Troponin elevation in coronary vs. non- coronary disease. European Heart Journal, 32(4). https://doi.org/10.1093/eurheartj/ehq456 Agüero-Torres, H., von Strauss, E., Viitanen, M., Winblad, B., & Fratiglioni, L. (2001). Institutionalization in the elderly: The role of chronic diseases and dementia. Cross-sectional and longitudinal data from a population-based study. Journal of Clinical Epidemiology, 54(8), 795– 801. https://doi.org/10.1016/s0895-4356(00)00371-1 Aimo, A., Januzzi, J. L., Vergaro, G. G. V., Ripoli, A., Latini, R., Masson, S., et al. (2018). Prognostic value of high-sensitivity troponin T in chronic heart failure an individual patient data meta-analysis. Circulation, 137(3), 286–297. https://doi.org/10.1161/circulationaha.117.031560 Airaksinen, K. E. J., Aalto, R., Hellman, T., Vasankari, T., Lahtinen, A., & Wittfooth, S. (2022). Novel Troponin Fragmentation Assay to Discriminate Between Troponin Elevations in Acute Myocardial Infarction and End-Stage Renal Disease. Circulation, 146(18), 1408–1410. https://doi.org/10.1161/circulationaha.122.060845 Ambrose, J. A., & Barua, R. S. (2004). The pathophysiology of cigarette smoking and cardiovascular disease: An update. Journal of the American College of Cardiology, 43(10), 1731–1737. https://doi.org/10.1016/j.jacc.2003.12.047 Antelmi, I., De Paula, R. S., Shinzato, A. R., Peres, C. A., Mansur, A. J., & Grupi, C. J. (2004). Influence of age, gender, body mass index, and functional capacity on heart rate variability in a cohort of subjects without heart disease. American Journal of Cardiology, 93(3), 381–385. https://doi.org/10.1016/j.amjcard.2003.09.065 Askin, L., Tanriverdi, O., & Turkmen, S. (2020). Clinical importance of high- sensitivity troponin T in patients without coronary artery disease. Northern Clinics of Istanbul, 7(3), 305–310. https://doi.org/10.14744/nci.2019.71135 Bech, A. P., Wetzels, J. F. M., & Nijenhuis, T. (2017). Reference values of renal tubular function tests are dependent on age and kidney function. Physiological Reports, 5(23). https://doi.org/10.14814/phy2.13542 Bergmann, O., Bhardwaj, R. D., Bernard, S., Zdunek, S., Barnabé-Heide, F., Walsh, S., Zupicich, J., Alkass, K., Buchholz, B. A., Druid, H., Jovinge, S., & Frisén, J. (2009). Evidence for cardiomyocyte renewal in humans. Science, 324(5923), 98-102. https://doi.org/10.1126/science.1164680 Elisa Heikkilä 74 Bertolotti, M., Lancellotti, G., & Mussi, C. (2024). Changes in cholesterol homeostasis associated with aging and with age-related conditions: pathophysiological and clinical implications. Journal Of Gerontology and Geriatrics, 72(4), 32–42. https://doi.org/10.36150/2499-6564-N637 Bharucha, A. J., Pandav, R., Shen, C., Dodge, H. H., & Ganguli, M. (2004). Predictors of Nursing Facility Admission: A 12-Year Epidemiological Study in the United States. Journal of the American Geriatrics Society, 52(3), 434–439. https://doi.org/10.1111/j.1532- 5415.2004.52118.x Blodgett, J. M., Theou, O., Howlett, S. E., Wu, F. C. W., & Rockwood, K. (2016). A frailty index based on laboratory deficits in community-dwelling men predicted their risk of adverse health outcomes. Age and Ageing, 45(4), 463–468. https://doi.org/10.1093/ageing/afw054 Blodgett, J. M., Theou, O., Howlett, S. E., & Rockwood, K. (2017). A frailty index from common clinical and laboratory tests predicts increased risk of death across the life course. GeroScience, 39(4), 447–455. https://doi.org/10.1007/s11357-017-9993-7 Blodgett, J. M., Theou, O., Mitnitski, A., Howlett, S. E., & Rockwood, K. (2019). Associations between a laboratory frailty index and adverse health outcomes across age and sex. Aging Medicine, 2(1), 11–17. https://doi.org/10.1002/agm2.12055 Bravell, M. E., Berg, S., Malmberg, B., & Sundström, G. (2009). Sooner or later? A study of institutionalization in late life. Aging Clinical and Experimental Research, 21(4–5), 329–337. https://doi.org/10.1007/bf03324923 Bulló, M., Garcia-Aloy, M., Martínez-González, M. A., Corella, D., Fernández-Ballart, J. D., Fiol, M., et al. (2011). Association between a healthy lifestyle and general obesity and abdominal obesity in an elderly population at high cardiovascular risk. Preventive Medicine, 53(3), 155–161. https://doi.org/10.1016/j.ypmed.2011.06.008 Cardinaels, E. P. M., Daamen, M. A. M. J., Bekers, O., ten Kate, J., Niens, M., van Suijlen, J. D. E., et al. (2015). Clinical Interpretation of Elevated Concentrations of Cardiac Troponin T, but Not Troponin I, in Nursing Home Residents. Journal of the American Medical Directors Association, 16(10), 884–891. https://doi.org/10.1016/J.jamda.2015.06.026 Cardinaels, E. P. M., Mingels, A. M. A., Van Rooij, T., Collinson, P. O., Prinzen, F. W., & Van Dieijen- Visser, M. P. (2013). Time-dependent degradation pattern of cardiac troponin T following myocardial infarction. Clinical Chemistry, 59(7), 1083–1090. https://doi.org/10.1373/clinchem.2012.200543 Chaulin, A. M. (2022). The Importance of Cardiac Troponin Metabolism in the Laboratory Diagnosis of Myocardial Infarction (Comprehensive Review). BioMed Research International. https://doi.org/10.1155/2022/6454467 Clegg, A., Young, J., Iliffe, S., Rikkert, M. O., & Rockwood, K. (2013). Frailty in elderly people. The Lancet, 381(9868), 752–762. https://doi.org/10.1016/S0140-6736(12)62167-9 Clegg, A., Rogers, L., & Young, J. (2015). Diagnostic test accuracy of simple instruments for identifying frailty in community-dwelling older people: a systematic review. Age and Ageing, 44(1), 148–152. https://doi.org/10.1093/ageing/afu157 Clegg, A., Bates, C., Young, J., Ryan, R., Nichols, L., Teale, E. A., et al. (2016). Development and validation of an electronic frailty index using routine primary care electronic health record data. Age and Ageing, 45(3), 353–360. https://doi.org/10.1093/ageing/afw039 Clinical and Laboratory Standards Institute (2010). EP28-A3c Defining, Establishing, and Verifying Reference Intervals in the Clinical Laboratory; Approved Guideline (3rd edition). Cohen, C. I., Benyaminov, R., Rahman, M., Ngu, D., & Reinhardt, M. (2023). Frailty: A Multidimensional Biopsychosocial Syndrome. Medical Clinics of North America, 107(1), 183– 197. https://doi.org/10.1016/j.mcna.2022.04.006 Collerton, J., Martin-Ruiz, C., Davies, K., Hilkens, C. M., Isaacs, J., Kolenda, C., Parker, C., Dunn, M., Catt, M., Jagger, C., von Zglinicki, T., & Kirkwood, T. B. L. (2012). Frailty and the role of inflammation, immunosenescence and cellular ageing in the very old: cross-sectional findings from the Newcastle 85+ Study. Mechanisms of Ageing and Development, 133(6), 456–466. https://doi.org/10.1016/j.mad.2012.05.005 Collett, D. (2014). Modelling Survival Data in Medical Research (3rd edition). Chapman and Hall/CRC. pp. 17- 35. References 75 Collins, A. J., Pitt, B., Reaven, N., Funk, S., Mcgaughey, K., Wilson, D., & Bushinsky, D. A. (2017). Association of Serum Potassium with All-Cause Mortality in Patients with and without Heart Failure, Chronic Kidney Disease, and/or Diabetes. American Journal of Nephrology, 46(3), 213– 221. https://doi.org/10.1159/000479802 Conroy, R. M., Pyörälä K., Fitzgerald, A. P., Sans S., Menotti A., De Backer G., et al. on behalf of the SCORE project group (2003). Estimation of ten-year risk of fatal cardiovascular disease in Europe: the SCORE project. European Heart Journal, 24(11), 987–1003. https://doi.org/10.1016/S0195-668x(03)00114-3 Corona, G., Giuliani, C., Parenti, G., Norello, D., Verbalis, J. G., Forti, G., Maggi, M., & Peri, A. (2013). Moderate hyponatremia is associated with increased risk of mortality: Evidence from a meta-analysis. PLoS ONE, 8(12). https://doi.org/10.1371/journal.pone.0080451 Costello-Boerrigter, L. C., Boerrigter, G., Redfield, M. M., Rodeheffer, R. J., Urban, L. H., Mahoney, D. W., et al. (2006). Amino-Terminal Pro-B-Type Natriuretic Peptide and B-Type Natriuretic Peptide in the General Community Determinants and Detection of Left Ventricular Dysfunction. Journal of the American College of Cardiology, 47(2). https://doi.org/10.1016/j.jacc.2005.09.025 Culleton, B. F., Manns, B. J., Zhang, J., Tonelli, M., Klarenbach, S., & Hemmelgarn, B. R. (2006). Impact of anemia on hospitalization and mortality in older adults. Blood, 107(10), 3841–3846. https://doi.org/10.1182/blood-2005-10-4308 D’Agostino, R. B., Vasan, R. S., Pencina, M. J., Wolf, P. A., Cobain, M., Massaro, J. M., & Kannel, W. B. (2008). General cardiovascular risk profile for use in primary care: The Framingham heart study. Circulation, 117(6), 743–753. https://doi.org/10.1161/circulationaha.107.699579 Daniels, L. B., Clopton, P., DeFilippi, C. R., Sanchez, O. A., Bahrami, H., Lima, J. A. C., et al. (2015). Serial measurement of N-terminal pro-B-type natriuretic peptide and cardiac troponin T for cardiovascular disease risk assessment in the Multi-Ethnic Study of Atherosclerosis (MESA). American Heart Journal, 170(6), 1170. https://doi.org/10.1016/j.ahj.2015.09.010 de Lemos, J. A., Drazner, M. H., Omland, T., Ayers, C. R., Khera, A., Rohatgi, A., et al. (2010). Association of troponin T detected with a highly sensitive assay and cardiac structure and mortality risk in the general population. JAMA, 304(22), 2503–2512. https://doi.org/10.1001/jama.2010.1768 De Maeyer, R. P. H., & Chambers, E. S. (2021). The impact of ageing on monocytes and macrophages. Immunology Letters, 230, 1–10. https://doi.org/10.1016/j.imlet.2020.12.003 de Vries, N. M., Staal, J. B., van Ravensberg, C. D., Hobbelen, J. S. M., Olde Rikkert, M. G. M., & Nijhuis-van der Sanden, M. W. G. (2011). Outcome instruments to measure frailty: a systematic review. Ageing Research Reviews, 10(1), 104–114. https://doi.org/10.1016/j.arr.2010.09.001 deFilippi, C. R., Christenson, R. H., Gottdiener, J. S., Kop, W. J., & Seliger, S. L. (2010). Dynamic Cardiovascular Risk Assessment in Elderly People. The Role of Repeated N-Terminal Pro-B- Type Natriuretic Peptide Testing. Journal of the American College of Cardiology, 55(5), 441– 450. https://doi.org/10.1016/j.jacc.2009.07.069 Devereaux, P. J., Chan, M. T. V., Alonso-Coello, P., Walsh, M., Berwanger, O., Villar, J. C., et al. (2012). Association Between Postoperative Troponin Levels and 30-Day Mortality Among Patients Undergoing Noncardiac Surgery. JAMA, 307(21), 2295–2304. https://doi.org/10.1001/jama.2012.5502 Dhana, K., Koolhaas, C. M., Van Rossum, E. F. C., Ikram, M. A., Hofman, A., Kavousi, M., & Franco, O. H. (2016). Metabolically Healthy Obesity and the Risk of Cardiovascular Disease in the Elderly Population. PLoS One, 11(4), e0154273. https://doi.org/10.1371/journal.pone.0154273 DiNicolantonio, J. J., & McCarty, M. F. (2018). Is interleukin-6 the link between low LDL cholesterol and increased non-cardiovascular mortality in the elderly? Open Heart, 5(1), e000789. https://doi.org/10.1136/openhrt-2018-000789 Dlima, S. D., Harris, D., Aminu, A. Q., Hall, A., Todd, C., & Vardy, E. R. (2025). Frailty indices based on routinely collected data: a scoping review. The Journal of Frailty & Aging, 14(3), 100047. https://doi.org/10.1016/j.tjfa.2025.100047 Duntas, L. H. (2021). Aging and the hypothalamic-pituitary-thyroid axis. Vitamins and Hormones, 115, 1–14. https://doi.org/10.1016/bs.vh.2020.12.001 Elisa Heikkilä 76 Eggers, K. M., Lind, L., Venge, P., & Lindahl, B. (2009). Will the Universal Definition of Myocardial Infarction Criteria Result in an Overdiagnosis of Myocardial Infarction? The American Journal of Cardiology, 103(5), 588–591. https://doi.org/10.1016/j.amjcard.2008.11.007 Ellervik, C., Marott, J. L., Tybjærg-Hansen, A., Schnohr, P., & Nordestgaard, B. G. (2014). Total and Cause-Specific Mortality by Moderately and Markedly Increased Ferritin Concentrations: General Population Study and Metaanalysis. Clinical Chemistry, 60(11), 1419–1428. https://doi.org/10.1373/clinchem.2014.229013 Ellis, H. L., Wan, B., Yeung, M., Rather, A., Mannan, I., Bond, C., Harvey, C., Raja, N., Dutey-Magni, P., Rockwood, K., Davis, D., & Searle, S. D. (2020). Complementing chronic frailty assessment at hospital admission with an electronic frailty index (FI-Laboratory) comprising routine blood test results. CMAJ: Canadian Medical Association Journal, 192(1), E3–E8. https://doi.org/10.1503/cmaj.190952 Epstein, F. H., Levin, E. R., Gardner, D. G., & Samson, W. K. (1998). Natriuretic Peptides. 339(5), 321–328. https://doi.org/10.1056/NEJM199807303390507 Félix-Redondo, F. J., Grau, M., & Fernández-Bergés, D. (2013). Cholesterol and cardiovascular disease in the elderly. Facts and gaps. Aging and Disease, 4(3), 154–169. PMCID: PMC3660125. Finnish Institute for Health and Welfare. (2022). THL's Morbidity Index 2019 Large regional differences in morbidity. https://www.julkari.fi/server/api/core/bitstreams/a32e253d-e97b-459c- 88ee-24461013a1ff /content Finnish Institute for Health and Welfare. (2025). Sosiaalihuollon laitos- ja asumispalvelut 2024. https://thl.fi/tilastot-ja-data/tilastot-aiheittain/ikaantyneet/sosiaalihuollon-laitos-ja- asumispalvelut Fleg, J. L., Aronow, W. S., & Frishman, W. H. (2011). Cardiovascular drug therapy in the elderly: Benefits and challenges. Nature Reviews Cardiology, 8(1), 13–28. https://doi.org/10.1038/nrcardio.2010.162 Fradley, M. G., Larson, M. G., Cheng, S., McCabe, E., Coglianese, E., Shah, R. V., Levy, D., Vasan, R. S., & Wang, T. J. (2011). Reference limits for N-terminal-pro-B-type natriuretic peptide in healthy individuals (from the Framingham Heart Study). The American Journal of Cardiology, 108(9), 1341–1345. https://doi.org/10.1016/j.amjcard.2011.06.057 Franzini, M., Lorenzoni, V., Masotti, S., Prontera, C., Chiappino, D., Latta, D. et al. (2015). The calculation of the cardiac troponin T 99th percentile of the reference population is affected by age, gender, and population selection: A multicenter study in Italy. Clinica Chimica Acta, 438, 376–381. https://doi.org/10.1016/j.cca.2014.09.010 Fried, L. P., Tangen, C. M., Walston, J., Newman, A. B., Hirsch, C., Gottdiener, J., Seeman, T., Tracy, R., Kop, W. J., Burke, G., & McBurnie, M. A. (2001). Frailty in older adults: evidence for a phenotype. J Gerontol A Biol Sci Med Sci., 56(3), M146–M156. https://doi.org/10.1093/gerona/56.3.m146 Galasko, G. I. W., Lahiri, A., Barnes, S. C., Collinson, P., & Senior, R. (2005). What is the normal range for N-terminal pro-brain natriuretic peptide? How well does this normal range screen for cardiovascular disease? European Heart Journal, 26(21), 2269–2276. https://doi.org/10.1093/eurheartj/ehi410 Garmany, A., Yamada, S., & Terzic, A. (2021). Longevity leap: mind the healthspan gap. NPJ Regenerative Medicine, 6(1), 57. https://doi.org/10.1038/s41536-021-00169-5 Gauthier, B. R., Sola-García, A., Cáliz-Molina, M. Á., Lorenzo, P. I., Cobo-Vuilleumier, N., Capilla- González, V., & Martin-Montalvo, A. (2020). Thyroid hormones in diabetes, cancer, and aging. Aging Cell, 19(11), e13260. https://doi.org/10.1111/acel.13260 Gazoti Debessa, C. R., Mesiano Maifrino, L. B., & Rodrigues de Souza, R. (2001). Age related changes of the collagen network of the human heart. Mechanisms of Ageing and Development, 122(10), 1049–1058. https://doi.org/10.1016/s0047-6374(01)00238-x Gekle, M. (2017). Kidney and aging — A narrative review. Experimental Gerontology, 87, 153–155. https://doi.org/10.1016/j.exger.2016.03.013 Gencer, B., Marston, N. A., Im, K., Cannon, C. P., Sever, P., Keech, A., Braunwald, E., Giugliano, R. P., & Sabatine, M. S. (2020). Efficacy and safety of lowering LDL cholesterol in older patients: References 77 a systematic review and meta-analysis of randomised controlled trials. The Lancet (British Edition), 396(10263), 1637–1643. https://doi.org/10.1016/s0140-6736(20)32332-1 Giovannini, S., Onder, G., Liperoti, R., Russo, A., Carter, C., Capoluongo, E., Pahor, M., Bernabei, R., & Landi, F. (2011). Interleukin-6, C-reactive protein, and tumor necrosis factor-alpha as predictors of mortality in frail, community-living elderly individuals. Journal of the American Geriatrics Society, 59(9), 1679–1685. https://doi.org/10.1111/j.1532-5415.2011.03570.x Gnjidic, D., Stanaway, F. F., Cumming, R., Waite, L., Blyth, F., Naganathan, V., Handelsman, D. J., & Le Couteur, D. G. (2012). Mild cognitive impairment predicts institutionalization among older men: a population-based cohort study. PloS One, 7(9), e46061. https://doi.org/10.1371/journal.pone.0046061 Gonen, M. (2007). Analyzing Receiver Operating Characteristic Curves Using SAS. SAS Institute. Goodman, S. G., Steg, P. G., Eagle, K. A., Fox, K. A. A., López-Sendón, J., Montalescot, G., Budaj, A., Kennelly, B. M., Gore, J. M., Allegrone, J., Granger, C. B., & Gurfinkel, E. P. (2006). The diagnostic and prognostic impact of the redefinition of acute myocardial infarction: Lessons from the Global Registry of Acute Coronary Events (GRACE). American Heart Journal, 151(3), 654– 660. https://doi.org/10.1016/j.ahj.2005.05.014 Gordon, E. H., & Hubbard, R. E. (2020). Differences in frailty in older men and women. Medical Journal of Australia, 212(4), 183–188. https://doi.org/10.5694/mja2.50466 Gospodarczyk, A., Marczewski, K., Gospodarczyk, N., Widuch, M., Tkocz, M., & Zalejska-Fiolka, J. (2022). Homocysteine And Cardiovascular Disease - A Current Review. Wiadomosci Lekarskie, 75(11), 2862–2866. https://doi.org/10.36740/wlek202211224 Goudot, F.-X., Boukertouta, T., Lazureanu, A., MSadek, S., Martins-Meune, E., Chenevier-Gobeaux, C., & Meune, C. (2021). NT-proBNP use in old patients and the impact of the selected threshold value: a big data analysis. Clinical Chemistry and Laboratory Medicine, 59(12), 461–464. https://doi.org/10.1515/cclm-2021-0547 Gualandro, D. M., Puelacher, C., LuratiBuse, G., Lampart, A., Strunz, C., Cardozo, F. A., et al. (2018). Comparison of high-sensitivity cardiac troponin I and T for the prediction of cardiac complications after non-cardiac surgery. American Heart Journal, 203, 67–73. https://doi.org/10.1016/j.ahj.2018.06.012 Guralnik, J., Ershler, W., Artz, A., Lazo-Langner, A., Walston, J., Pahor, M., Ferrucci, L., & Evans, W. J. (2022). Unexplained anemia of aging: Etiology, health consequences, and diagnostic criteria. Journal of the American Geriatrics Society, 70(3), 891. https://doi.org/10.1111/jgs.17565 Hajek, A., Brettschneider, C., Lange, C., Posselt, T., Wiese, B., Steinmann, S., et al. (2015). Longitudinal Predictors of Institutionalization in Old Age. PloS One, 10(12). https://doi.org/10.1371/journal.pone.0144203 Hajek, A., Bock, J.-O., Saum, K.-U., Matschinger, H., Brenner, H., Holleczek, B., Haefeli, W. E., Heider, D., & König, H.-H. (2018). Frailty and healthcare costs-longitudinal results of a prospective cohort study. Age and Ageing, 47(2), 233–241. https://doi.org/10.1093/ageing/afx157 Hakeem, F. F., Maharani, A., Todd, C., & O’Neill, T. W. (2023). Development, validation and performance of laboratory frailty indices: A scoping review. Archives of Gerontology and Geriatrics, 111, 104995. https://doi.org/10.1016/j.archger.2023.104995 Halonen P, Raitanen J, Jämsen E, Enroth L, Jylhä M. (2019). Chronic conditions and multimorbidity in population aged 90 years and over: associations with mortality and long-term care admission. Age Ageing. 48(4):564-570. doi: 10.1093/ageing/afz019 Hao, Q., Sun, X., Yang, M., Dong, B., Dong, B., & Wei, Y. (2019). Prediction of mortality in Chinese very old people through the frailty index based on routine laboratory data. Scientific Reports, 9(1). https://doi.org/10.1038/s41598-018-36569-9 Harris, T. B., Ferrucci, L., Tracy, R. P., Chiara Corti, M., Wacholder, S., Ettinger, W. H., Heimovitz, H., Cohen, H. J., & Wallace, R. (1999). Associations of Elevated Interleukin-6 and C-Reactive Protein Levels with Mortality in the Elderly. Am J Med, 106, 506–512. https://doi.org/10.1016/s0002-9343(99)00066-2 Elisa Heikkilä 78 He, X., Jing, W., Zhu, R., Wang, Q., Yang, J., Tang, X., et al. (2025). Association of Reversible Frailty with All-Cause Mortality Risk in Community-Dwelling Older Adults and Analysis of Factors Affecting Frailty Reversal in Older Adults. Journal of the American Medical Directors Association, 26(5), Article 105527. https://doi.org/10.1016/j.jamda.2025.105527 Hilvo, M., Meikle, P. J., Pedersen, E. R., Tell, G. S., Dhar, I., Brenner, H., et al. (2020). Development and validation of a ceramide- and phospholipid-based cardiovascular risk estimation score for coronary artery disease patients. European Heart Journal, 41(3), 371–380. https://doi.org/10.1093/eurheartj/ehz387 Hoffmann, S., Wiben, A., Kruse, M., Jacobsen, K. K., Lembeck, M. A., & Holm, E. A. (2020). Predictive validity of PRISMA-7 as a screening instrument for frailty in a hospital setting. BMJ Open, 10(10). https://doi.org/10.1136/bmjopen-2020-038768 Holme, I., & Anderssen, S. A. (2015). Increases in physical activity is as important as smoking cessation for reduction in total mortality in elderly men: 12 years of follow-up of the Oslo II study. British Journal of Sports Medicine, 49(11), 743–748. https://doi.org/10.1136/bjsports-2014-094522 Horn, P. S., & Pesce, A. J. (2003). Reference intervals: an update. Clinica Chimica Acta, 334(1–2), 5– 23. https://doi.org/10.1016/s0009-8981(03)00133-5 Howlett, S. E., & Rockwood, K. (2013). New horizons in frailty: Ageing and the deficit-scaling problem. Age and Ageing, 42(4), 416–423. https://doi.org/10.1093/ageing/aft059 Howlett, S. E., Rockwood, M. R. H., Mitnitski, A., & Rockwood, K. (2014). Standard laboratory tests to identify older adults at increased risk of death. BMC Medicine, 12, 171. https://doi.org/10.1186/s12916-014-0171-9 Hupin, D., Roche, F., Gremeaux, V., Chatard, J. C., Oriol, M., Gaspoz, J. M., Barthélémy, J. C., & Edouard, P. (2015). Even a low-dose of moderate-to-vigorous physical activity reduces mortality by 22% in adults aged ≥60 years: A systematic review and meta-analysis. British Journal of Sports Medicine, 49(19) 1262–1267. https://doi.org/10.1136/bjsports-2014-094306 Jäger, J., Sieber, C. C., Gaßmann, K.-G., & Ritt, M. (2019). Changes of a frailty index based on common blood and urine tests during a hospital stay on geriatric wards predict 6-month and 1-year mortality in older people. Clinical Interventions in Aging, 14, 473–484. https://doi.org/10.2147/cia.s191117 Jansen, H. I., Dirks, N. F., Hillebrand, J. J., ten Boekel, E., Brinkman, J. W., Buijs, et al. (2024). Age- Specific Reference Intervals for Thyroid-Stimulating Hormones and Free Thyroxine to Optimize Diagnosis of Thyroid Disease. Thyroid: Official Journal of the American Thyroid Association, 34(11). https://doi.org/10.1089/thy.2024.0346 Januzzi, J. L., Van Kimmenade, R., Lainchbury, J., Bayes-Genis, A., Ordonez-Llanos, J., Santalo-Bel, M., Pinto, Y. M., & Richards, M. (2006). NT-proBNP testing for diagnosis and short-term prognosis in acute destabilized heart failure: an international pooled analysis of 1256 patients: the International Collaborative of NT-proBNP Study. European Heart Journal, 27(3), 330–337. https://doi.org/10.1093/eurheartj/ehi631 Jousilahti P, Laatikainen T, Peltonen M, Borodulin K, Männistö S, Jula A, et al. (2016). Primary prevention and risk factor reduction in coronary heart disease mortality among working aged men and women in eastern Finland over 40 years: population based observational study. BMJ 2352:i721. doi:10.1136/bmj.i721 Kaikkonen, J. E., Mikkilä, V., Magnussen, C. G., Juonala, M., Viikari, J. S. A., & Raitakari, O. T. (2013). Does childhood nutrition influence adult cardiovascular disease risk?-Insights from the Young Finns Study. Annals of Medicine, 45(2), 120–128. https://doi.org/10.3109/07853890.2012.671537 Kaptoge, S., Seshasai, S. R. K., Sun, L., Walker, M., Bolton, T., Spackman, S., et al. (2023). Life expectancy associated with different ages at diagnosis of type 2 diabetes in high-income countries: 23 million person-years of observation. The Lancet Diabetes & Endocrinology, 11(10), 731–742. https://doi.org/10.1016/S2213-8587(23)00223-1 Katajamäki, T. T., Koivula, M.-K., Hilvo, M., Lääperi, M. T. A., Salminen, M. J., Viljanen, A. M., Heikkilä, E. T. M., et al. (2022). Ceramides and Phosphatidylcholines Associate with Cardiovascular Diseases in the Elderly. Clinical Chemistry, 68(12), 1502–1508. https://doi.org/10.1093/clinchem/hvac158 References 79 Katajamäki, T. T., Koivula, M.-K., Salminen, M. J., Vahlberg, T., Heikkilä, E. T. M., Viljanen, A. M., Löppönen, M. K., Isoaho, R. E., Kivelä, S.-L., Viitanen, M., Viikari, J., Viikari, L., Pulkki, K. J., & Irjala, K. M. (2025). Small dense low-density lipoprotein as biomarker in the elderly. Clinical Biochemistry, 137, 110916. https://doi.org/10.1016/j.clinbiochem.2025.110916 Kauko, A., Engler, D., Niiranen, T., Ortega-Alonso, A., & Schnabel, R. B. (2024). Increased risk of dementia differs across cardiovascular diseases and types of dementia – Data from a nationwide study. Journal of Internal Medicine, 295(2), 196–205. https://doi.org/10.1111/joim.13733 Kirkwood, T. B. L. (2011). Systems biology of ageing and longevity. Philosophical Transactions of the Royal Society B: Biological Sciences, 366(1561), 64–70. https://doi.org/10.1098/rstb.2010.0275 Kistorp, C., Raymond, I., Pedersen, F., Gustafsson, F., Faber, J., & Hildebrandt, P. (2005). N-Terminal Pro-Brain Natriuretic Peptide, C-Reactive Protein, and Urinary Albumin Levels as Predictors of Mortality and Cardiovascular Events in Older Adults. JAMA, 293(13), 1609–1616. https://doi.org/10.1001/jama.293.13.1609 Klausen, H. H., Petersen, J., Bandholm, T., Juul-Larsen, H. G., Tavenier, J., Eugen-Olsen, J., & Andersen, O. (2017). Association between routine laboratory tests and long-term mortality among acutely admitted older medical patients: a cohort study. BMC Geriatrics, 17(1), 62. https://doi.org/10.1186/s12877-017-0434-3 Kovesdy, C. P., Lott, E. H., Lu, J. L., Malakauskas, S. M., Ma, J. Z., Molnar, M. Z., & Kalantar-Zadeh, K. (2012). Hyponatremia, hypernatremia, and mortality in patients with chronic kidney disease with and without congestive heart failure. Circulation, 125(5), 677–684. https://doi.org/10.1161/circulationaha.111.065391 Krogager, M. L., Torp-Pedersen, C., Mortensen, R. N., Køber, L., Gislason, G., Søgaard, P., & Aasbjerg, K. (2017). Short-term mortality risk of serum potassium levels in hypertension: a retrospective analysis of nationwide registry data. European Heart Journal, 38(2), 104–112. https://doi.org/10.1093/eurheartj/ehw129 Kulminski, A., Yashin, A., Ukraintseva, S., Akushevich, I., Arbeev, K., Land, K., & Manton, K. (2006). Accumulation of health disorders as a systemic measure of aging: Findings from the NLTCS data. Mechanisms of Ageing and Development, 127(11), 840–848. https://doi.org/10.1016/j.mad.2006.08.005 Lakatta, E. G. (2002). Age-associated Cardiovascular Changes in Health: Impact on Cardiovascular Disease in Older Persons. Heart Fail Rev. 7 29-49. https://doi.org/10.1023/A:1013797722156 Lakatta, E. G., & Levy, D. (2003). Arterial and cardiac aging: Major shareholders in cardiovascular disease enterprises: Part I: Aging arteries: A “set up” for vascular disease. Circulation, 107(1), 139–146. https://doi.org/10.1161/01.cir.0000048892.83521.58 Lam, T. H., Schooling, C. M., Li, Z. B, Ho, S. Y., Chan, W. M., Ho, K. S., et al. (2007). Smoking and mortality in the oldest-old, evidence from a prospective cohort of 56,000 Hong Kong Chinese. Journal of the American Geriatrics Society, 55(12), 2090–2091. https://doi.org/10.1111/j.1532- 5415.2007.01441.x Latini, R., Masson, S., Anand, I. S., Missov, E., Carlson, M., Vago, T., Angelici, L., Barlera, S., Parrinello, G., Maggioni, A. P., Tognoni, G., & Cohn, J. N. (2007). Prognostic value of very low plasma concentrations of troponin T in patients with stable chronic heart failure. Circulation, 116(11), 1242–1249. https://doi.org/10.1161/circulationaha.106.655076 Lear, S. A., Hu, W., Rangarajan, S., Gasevic, D., Leong, D., Iqbal, R., et al. (2017). The effect of physical activity on mortality and cardiovascular disease in 130 000 people from 17 high- income, middle-income, and low-income countries: the PURE study. The Lancet, 390(10113), 2643–2654. https://doi.org/10.1016/s0140-6736(17)31634-3 Leong, D. P., Diaz, Yusuf, R., Iqbal, R., Avezum, A., Yusufali, A., Rosengren, A., et al. (2025). The burden of cardiovascular events according to cardiovascular risk profile in adults from high- income, middle-income, and low-income countries (PURE): a cohort study. The Lancet global health, 13(8), e1406-e1414 https://doi.org/10.1016/s2214-109x(25)00155-x Letois, F., Mura, T., Scali, J., Gutierrez, L. A., Féart, C., & Berr, C. (2016). Nutrition and mortality in the elderly over 10 years of follow-up: the Three-City study. British Journal of Nutrition, 116(5), 882–889. https://doi.org/10.1017/S000711451600266x Elisa Heikkilä 80 Lewington, S., Clarke, R., Qizilbash, N., Peto, R., & Collins, R. (2002). Age-specific relevance of usual blood pressure to vascular mortality: A meta-analysis of individual data for one million adults in 61 prospective studies. Lancet, 360(9349), 1903–1913. https://doi.org/10.1016/s0140- 6736(02)11911-8 Li, W. J., Chen, X. M., Nie, X. Y., Zhang, J., Cheng, Y. J., Lin, X. X., & Wu, S. H. (2015). Cardiac troponin and C-reactive protein for predicting all-cause and cardiovascular mortality in patients with chronic kidney disease: A meta-analysis. Clinics, 70(4), 301–311. https://doi.org/10.6061/clinics/2015(04)14 Li, Y., Zhong, X., Cheng, G., Zhao, C., Zhang, L., Hong, Y., Wan, Q., He, R., & Wang, Z. (2017). Hs- CRP and all-cause, cardiovascular, and cancer mortality risk: A meta-analysis. Atherosclerosis, 259, 75–82. https://doi.org/10.1016/j.atherosclerosis.2017.02.003 Liu, L., Cheng, Y. T., Xu, A., & Cheung, B. M. Y. (2023). Association between high sensitivity cardiac troponin and mortality risk in the non-diabetic population: findings from the National Health and Nutrition Examination Survey. Cardiovascular Diabetology, 22(1), 1–10. https://doi.org/10.1186/s12933-023-02003-2 Löppönen, M., Räihä, I., Isoaho, R., Vahlberg, T., & Kivelä, S.-L. (2003). Diagnosing cognitive impairment and dementia in primary health care -- a more active approach is needed. Age and Ageing, 32(6), 606–612. https://doi.org/10.1093/ageing/afg097 Lorgis, L., Zeller, M., Dentan, G., Sicard, P., Buffet, P., L’Huillier, I., Beer, J. C., Vincent-Martin, M., Makki, H., Gambert, P., & Cottin, Y. (2009). Prognostic value of N-terminal pro-brain natriuretic peptide in elderly people with acute myocardial infarction: prospective observational study. BMJ (Clinical Research Ed.), 338(7704), 1195. https://doi.org/10.1136/bmj.b1605 Luppa, M., Luck, T., Weyerer, S., König, H.-H., Brähler, E., & Riedel-Heller, S. G. (2010). Prediction of institutionalization in the elderly. A systematic review. Age and Ageing, 39(1), 31–38. https://doi.org/10.1093/ageing/afp202 MacMahon, S., Duffy, S., Rodgers, A., Tominaga, S., Chambless, L., De Backer, G., et al. (2007). Blood cholesterol and vascular mortality by age, sex, and blood pressure: a meta-analysis of individual data from 61 prospective studies with 55 000 vascular deaths. The Lancet (British Edition), 370(9602), 1829–1839. https://doi.org/10.1016/s0140-6736(07)61778-4 Magnussen, C., Ojeda, F. M., Leong, D. P., Alegre-Diaz, J., Amouyel, P., Aviles-Santa, L., et al. (2023). Global Impact of Modifiable Risk Factors on Cardiovascular Disease and Mortality. The New England Journal of Medicine, 389(14), 1273. https://doi.org/10.1056/nejmoa2206916 Mahroum, N., Alghory, A., Kiyak, Z., Alwani, A., Seida, R., Alrais, M., & Shoenfeld, Y. (2022). Ferritin – from iron, through inflammation and autoimmunity, to COVID-19. Journal of Autoimmunity, 126, 102778. https://doi.org/10.1016/j.jaut.2021.102778 Mair, J., Lindahl, B., Müller, C., Giannitsis, E., Huber, K., Möckel, M., Plebani, M., Thygesen, K., & Jaffe, A. S. (2018). What to do when you question cardiac troponin values. European Heart Journal. Acute Cardiovascular Care, 7(6), 577–586. https://doi.org/10.1177/2048872617708973 Makino, K., Lee, S., Bae, S., Chiba, I., Harada, K., Katayama, O., Shinkai, Y., & Shimada, H. (2021). Absolute Cardiovascular Disease Risk Assessed in Old Age Predicts Disability and Mortality: A Retrospective Cohort Study of Community-Dwelling Older Adults. Journal of the American Heart Association, 10(24), 22004. https://doi.org/10.1161/jaha.121.022004 Mammen, J. S. (2019). Interpreting elevated TSH in older adults. Current Opinion in Endocrine and Metabolic Research, 5, 68–73. https://doi.org/10.1016/j.coemr.2019.04.001 Martinez-Rumayor, A., Richards, A. M., Burnett, J. C., & Januzzi, J. L. (2008). Biology of the Natriuretic Peptides. The American Journal of Cardiology, 101(3), S3–S8. https://doi.org/10.1016/j.amjcard.2007.11.012 McCarthy, C. P., Wasfy, J. H., & Januzzi, J. L. (2024). Is Myocardial Infarction Overdiagnosed? JAMA : The Journal of the American Medical Association, 331(19), 1623–1624. https://doi.org/10.1001/jama.2024.5235 Menacer, S., Claessens, Y. E., Meune, C., Elfassi, Y., Wakim, C., Gauthier, L., Fortun, M., Goudot, F. X., Dehoux, M., Lefèvre, G., & Chenevier-Gobeaux, C. (2013). Reference range values of troponin measured by sensitive assays in elderly patients without any cardiac signs/symptoms. Clinica Chimica Acta, 417, 45–47. https://doi.org/10.1016/j.cca.2012.11.031 References 81 Millán-Calenti, J. C., Tubío, J., Pita-Fernández, S., González-Abraldes, I., Lorenzo, T., Fernández- Arruty, T., & Maseda, A. (2010). Prevalence of functional disability in activities of daily living (ADL), instrumental activities of daily living (IADL) and associated factors, as predictors of morbidity and mortality. Archives of Gerontology and Geriatrics, 50(3), 306–310. https://doi.org/10.1016/j.archger.2009.04.017 Mills, K. T., Bundy, J. D., Kelly, T. N., Reed, J. E., Kearney, P. M., Reynolds, K., Chen, J., & He, J. (2016). Global Disparities of Hypertension Prevalence and Control: A Systematic Analysis of Population-Based Studies From 90 Countries. Circulation, 134(6), 441–450. https://doi.org/10.1161/circulationaha.115.018912 Mingels, A. M. A., Cardinaels, E. P. M., Broers, N. J. H., Van Sleeuwen, A., Streng, A. S., Van Dieijen- Visser, M. P., Kooman, J. P., & Bekers, O. (2017). Cardiac Troponin T: Smaller Molecules in Patients with End-Stage Renal Disease than after Onset of Acute Myocardial Infarction. Clinical Chemistry, 63(3), 683–690. https://doi.org/10.1373/clinchem.2016.261644 Mitnitski, A. B., Mogilner, A. J., & Rockwood, K. (2001). Accumulation of Deficits as a Proxy Measure of Aging. 1, 323–336. https://doi.org/10.1100/tsw.2001.58 Mitnitski, A. B., Graham, J. E., Mogilner, A. J., & Rockwood, K. (2002). Frailty, fitness and late-life mortality in relation to chronological and biological age. BMC Geriatrics, 2, 1. https://doi.org/10.1186/1471-2318-2-1 Mitnitski, A., Song, X., & Rockwood, K. (2012). Trajectories of changes over twelve years in the health status of Canadians from late middle age. Experimental Gerontology, 47(12), 893–899. https://doi.org/10.1016/j.exger.2012.06.015 Mitnitski, A., Song, X., & Rockwood, K. (2013). Assessing biological aging: The origin of deficit accumulation. Biogerontology, 14(6), 709–717. https://doi.org/10.1007/s10522-013-9446-3 Mitnitski, A., Collerton, J., Martin-Ruiz, C., Jagger, C., von Zglinicki, T., Rockwood, K., & Kirkwood, T. B. L. (2015). Age-related frailty and its association with biological markers of ageing. BMC Medicine, 13, 161. https://doi.org/10.1186/s12916-015-0400-x Mitnitski, A., & Rockwood, K. (2016). The rate of aging: the rate of deficit accumulation does not change over the adult life span. Biogerontology, 17(1), 199–204. https://doi.org/10.1007/s10522- 015-9583-y Mitrou, P., Raptis, S. A., & Dimitriadis, G. (2011). Thyroid disease in older people. Maturitas, 70(1), 5–9. https://doi.org/10.1016/j.maturitas.2011.05.016 Mohan, S., Gu, S., Parikh, A., & Radhakrishnan, J. (2013). Prevalence of hyponatremia and association with mortality: Results from NHANES. American Journal of Medicine, 126(12). https://doi.org/10.1016/j.amjmed.2013.07.021 Morley, J. E. (2004). The Metabolic Syndrome and Aging. The Journals of Gerontology: Series A, 59(2), M139–M142. https://doi.org/10.1093/gerona/59.2.m139 Morley, J. E., Vellas, B., van Kan, G. A., Anker, S. D., Bauer, J. M., Bernabei, R., et al. (2013). Frailty consensus: a call to action. Journal of the American Medical Directors Association, 14(6), 392– 397. https://doi.org/10.1016/j.jamda.2013.03.022 Moshkovits, Y., Chetrit, A., & Dankner, R. (2024). The association between frailty biomarkers and 20- year all-cause and cardiovascular mortality among community-dwelling older adults. Postgraduate Medicine, 136(6), 641–650. https://doi.org/10.1080/00325481.2024.2374703 Nadrowski, P., Chudek, J., Grodzicki, T., Mossakowska, M., Skrzypek, M., Wiecek, A., Zdrojewski, T., & Kozakiewicz, K. (2013). Plasma level of N-terminal pro brain natriuretic peptide (NT- proBNP) in elderly population in Poland — The PolSenior Study. Experimental Gerontology, 48(9), 852–857. https://doi.org/10.1016/j.exger.2013.05.060 Nanna, M. G., Navar, A. M., Wojdyla, D., & Peterson, E. D. (2019). The Association Between Low- Density Lipoprotein Cholesterol and Incident Atherosclerotic Cardiovascular Disease in Older Adults: Results From the National Institutes of Health Pooled Cohorts. Journal of the American Geriatrics Society, 67(12), 2560–2567. https://doi.org/10.1111/jgs.16123 National Guideline Centre (UK). (2018). Chronic Heart Failure in Adults: Diagnosis and Management. National Institute for Health and Care Excellence (NICE). Elisa Heikkilä 82 Ndrepepa, G., & Kastrati, A. (2019). Alanine aminotransferase—a marker of cardiovascular risk at high and low activity levels. Journal of Laboratory and Precision Medicine, 4(0). https://doi.org/10.21037/jlpm.2019.08.01 Neumann, J. T., de Lemos, J. A., Apple, F. S., & Leong, D. P. (2025). Cardiovascular biomarkers for risk stratification in primary prevention. European Heart Journal, 00, 1–21. https://doi.org/10.1093/eurheartj/ehaf517 Neumann, J. T., Twerenbold, R., Weimann, J., Ballantyne, C. M., Benjamin, E. J., Costanzo, et al. (2024). Prognostic Value of Cardiovascular Biomarkers in the Population. JAMA, 331(22), 1898-1909. https://doi.org/10.1001/jama.2024.5596 Ng, T. P., Feng, L., Nyunt, M. S. Z., Feng, L., Niti, M., Tan, B. Y., Chan, G., Khoo, S. A., Chan, S. M., Yap, P., & Yap, K. B. (2015). Nutritional, Physical, Cognitive, and Combination Interventions and Frailty Reversal Among Older Adults: A Randomized Controlled Trial. The American Journal of Medicine, 128(11), 1225-1236.e1. https://doi.org/10.1016/j.amjmed.2015.06.017 Ni, W., Zhang, M., Wang, X., Li, X., Wang, Q., Wang, Y., Chen, G., Shen, T., Fan, K., Yao, X., Sun, Y., Liu, C., & Xu, S. (2022). Age-specific serum thyrotropin reference range for the diagnosis of subclinical hypothyroidism and its association with lipid profiles in the elderly population. Scientific Reports, 12(1), 1–8. https://doi.org/10.1038/s41598-022-24182-w Nicholls, C., Li, H., Wang, J. Q., & Liu, J. P. (2011). Molecular regulation of telomerase activity in aging. Protein & Cell, 2(9), 726. https://doi.org/10.1007/s13238-011-1093-3 North, B. J., & Sinclair, D. A. (2012). The Intersection Between Aging and Cardiovascular Disease. Circulation Research, 110(8), 1097. https://doi.org/10.1161/circresaha.111.246876 Nuotio, J., Laitinen, T. T., Juonala, M. (2020). Mistä ero sepelvaltimotaudissa idän ja lännen välillä johtuu? Lääketieteellinen Aikakauskirja Duodecim 136(23):2583-9 Official Statistics Finland (2024). Causes of death. https://stat.fi/tilasto/ksyyt Official Statistics Finland (2025). Väestö ja väestöennuste ikäryhmittäin. https://stat.fi/tup/suoluk/suoluk_vaesto.html Okamoto, R., Ali, Y., Hashizume, R., Suzuki, N., & Ito, M. (2019). BNP as a Major Player in the Heart- Kidney Connection. International Journal of Molecular Sciences, 20(14). https://doi.org/10.3390/ijms20143581 Olivetti, G., Melissari, M., Capasso, J. M., & Anversa, P. (1991). Cardiomyopathy of the aging human heart. Myocyte loss and reactive cellular hypertrophy. Circulation Research, 68(6), 1560–1568. https://doi.org/10.1161/01.res.68.6.1560 Olivieri, F., Galeazzi, R., Giavarina, D., Testa, R., Abbatecola, A. M., Çeka, A., Tamburrini, P., Busco, F., Lazzarini, R., Monti, D., Franceschi, C., Procopio, A. D., & Antonicelli, R. (2012). Aged- related increase of high sensitive Troponin T and its implication in acute myocardial infarction diagnosis of elderly patients. Mechanisms of Ageing and Development, 133(5), 300–305. https://doi.org/10.1016/j.mad.2012.03.005 Organisation for Economic Co-operation and Development (n.d.). Elderly population. https://doi.org/10.1787/8d805ea1-en Parks, R. J., Fares, E., MacDonald, J. K., Ernst, M. C., Sinal, C. J., Rockwood, K., & Howlett, S. E. (2012). A procedure for creating a frailty index based on deficit accumulation in aging mice. Journals of Gerontology - Series A Biological Sciences and Medical Sciences, 67 A(3), 217–227. https://doi.org/10.1093/gerona/glr193 Pennells, L., Kaptoge, S., Østergaard, H. B., Read, S. H., Carinci, F., Franch-Nadal, J., et al. (2023). SCORE2-Diabetes: 10-year cardiovascular risk estimation in type 2 diabetes in Europe. European Heart Journal, 44(28), 2544–2556. https://doi.org/10.1093/eurheartj/ehad260 Perone, F., Bernardi, M., Spadafora, L., Betti, M., Cacciatore, S., Saia, F., Fogacci, F., Jaiswal, V., Asher, E., Paneni, F., De Rosa, S., Banach, M., Biondi Zoccai, G., & Sabouret, P. (2025). Non- Traditional Cardiovascular Risk Factors: Tailored Assessment and Clinical Implications. Journal of Cardiovascular Development and Disease 12(5), 171. https://doi.org/10.3390/jcdd12050171 Pialoux, T., Goyard, J., & Lesourd, B. (2012). Screening tools for frailty in primary health care: a systematic review. Geriatrics and gerontology international. 12(2), 189–197. https://doi.org/10.1111/j.1447-0594.2011.00797.x References 83 Ponikowski, P., Voors, A. A., Anker, S. D., Bueno, H., Cleland, J. G. F., Coats, A. J. S., et al. (2016). ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure. European Heart Journal, 37(27), 2129–2200. https://doi.org/10.1093/eurheartj/ehw128 Ravnskov, U., Diamond, D. M., Hama, R., Hamazaki, T., Hammarskjöld, B., Hynes, N., Kendrick, M., Langsjoen, P. H., Malhotra, A., Mascitelli, L., McCully, K. S., Ogushi, Y., Okuyama, H., Rosch, P. J., Schersten, T., Sultan, S., & Sundberg, R. (2016). Lack of an association or an inverse association between low-density-lipoprotein cholesterol and mortality in the elderly: A systematic review. BMJ Open, 6(6). https://doi.org/10.1136/bmjopen-2015-010401 Raymond, I., Groenning, B. A., Hildebrandt, P. R., Nilsson, J. C., Baumann, M., Trawinski, J., & Pedersen, F. (2003). The influence of age, sex and other variables on the plasma level of N- terminal pro brain natriuretic peptide in a large sample of the general population. Heart (British Cardiac Society), 89(7), 745–751. https://doi.org/10.1136/heart.89.7.745 Reichlin, T., Hochholzer, W., Bassetti, S., Steuer, S., Stelzig, C., Hartwiger, S., Biedert, S., Schaub, N., Buerge, C., Potocki, M., Noveanu, M., Breidthardt, T., Twerenbold, R., Winkler, K., Bingisser, R., & Mueller, C. (2009). Early Diagnosis of Myocardial Infarction with Sensitive Cardiac Troponin Assays. New England Journal of Medicine, 361(9), 858–867. https://doi.org/10.1056/nejmoa0900428 Reiter, M., Twerenbold, R., Reichlin, T., Haaf, P., Peter, F., Meissner, J., Hochholzer, W., Stelzig, C., Freese, M., Heinisch, C., Breidthardt, T., Freidank, H., Winkler, K., Campodarve, I., Gea, J., & Mueller, C. (2011). Early diagnosis of acute myocardial infarction in the elderly using more sensitive cardiac troponin assays. European Heart Journal, 32(11), 1379–1389. https://doi.org/10.1093/eurheartj/ehr033 Rifai, N., Horvath, A. R., & Wittwer, C. (2018). Tietz textbook of clinical chemistry and molecular diagnostics (Sixth edition.). Elsevier. Ritt, M., Jäger, J., Ritt, J. I., Sieber, C. C., & Gaßmann, K.-G. (2017). Operationalizing a frailty index using routine blood and urine tests. Clinical Interventions in Aging, 12, 1029–1040. https://doi.org/10.2147/cia.s131987 Rockwood, K., Howlett, S. E., MacKnight, C., Beattie, B. L., Bergman, H., Hébert, R., Hogan, D. B., Wolfson, C., & McDowell, I. (2004). Prevalence, attributes, and outcomes of fitness and frailty in community-dwelling older adults: report from the Canadian study of health and aging. The Journals of Gerontology. Series A, Biological Sciences and Medical Sciences, 59(12), 1310– 1317. https://doi.org/10.1093/gerona/59.12.1310 Rockwood, K., Andrew, M., & Mitnitski, A. (2007). A comparison of two approaches to measuring frailty in elderly people. The Journals of Gerontology. Series A, Biological Sciences and Medical Sciences, 62(7), 738–743. https://doi.org/10.1093/gerona/62.7.738 Rockwood, K., & Mitnitski, A. (2007). Frailty in relation to the accumulation of deficits. The journals of gerontology, 62(7) 722–727. https://doi.org/10.1093/gerona/62.7.722 Rockwood, K., McMillan, M., Mitnitski, A., & Howlett, S. E. (2015). A Frailty Index Based on Common Laboratory Tests in Comparison With a Clinical Frailty Index for Older Adults in Long-Term Care Facilities. Journal of the American Medical Directors Association, 16(10), 842–847. https://doi.org/10.1016/j.jamda.2015.03.027 Rodríguez-Castro, E., Hervella, P., López-Dequidt, I., Arias-Rivas, S., Santamaría-Cadavid, M., López-Loureiro, I., da Silva-Candal, A., Pérez-Mato, M., Sobrino, T., Campos, F., Castillo, J., Rodríguez-Yáñez, M., & Iglesias-Rey, R. (2020). NT-pro-BNP: A novel predictor of stroke risk after transient ischemic attack. International Journal of Cardiology, 298, 93–97. https://doi.org/10.1016/j.ijcard.2019.06.056 Rohde, L. E. P., Hennekens, C. H., & Ridker, P. M. (1999). Survey of C-reactive protein and cardiovascular risk factors in apparently healthy men. The American Journal of Cardiology, 84(9), 1018–1022. https://doi.org/10.1016/s0002-9149(99)00491-9 Rutenberg, A. D., Mitnitski, A. B., Farrell, S. G., & Rockwood, K. (2018). Unifying aging and frailty through complex dynamical networks. Experimental Gerontology, 107, 126–129. https://doi.org/10.1016/j.exger.2017.08.027 Salminen, M., Vire, J., Viikari, L., Vahlberg, T., Isoaho, H., Lehtonen, A., Viitanen, M., Arve, S., & Eloranta, S. (2017). Predictors of institutionalization among home-dwelling older Finnish Elisa Heikkilä 84 people: a 22-year follow-up study. Aging Clinical and Experimental Research, 29(3), 499–505. https://doi.org/10.1007/s40520-016-0722-3 Salminen, M., Viljanen, A., Eloranta, S., Viikari, P., Wuorela, M., Vahlberg, T., Isoaho, R., Kivelä, S.- L., Korhonen, P., Irjala, K., Löppönen, M., & Viikari, L. (2019). Frailty and mortality: an 18- year follow-up study among Finnish community-dwelling older people. Aging Clinical and Experimental Research. https://doi.org/10.1007/s40520-019-01383-4 Salminen, M., Laine, J., Vahlberg, T., Viikari, P., Wuorela, M., Viitanen, M., & Viikari, L. (2020). Factors associated with institutionalization among home-dwelling patients of Urgent Geriatric Outpatient Clinic: a 3-year follow-up study. European Geriatric Medicine. https://doi.org/10.1007/s41999-020-00338-7 Sapp, D. G., Cormier, B. M., Rockwood, K., Howlett, S. E., & Heinze, S. S. (2023). The frailty index based on laboratory test data as a tool to investigate the impact of frailty on health outcomes: a systematic review and meta-analysis. Age and Ageing, 52(1), afac309. https://doi.org/10.1093/ageing/afac309 Sarwar, N., Gao, P., Kondapally Seshasai, S. R., Gobin, R., Kaptoge, S., Di Angelantonio, E., Ingelsson, E., et al. (2010). Diabetes mellitus, fasting blood glucose concentration, and risk of vascular disease: a collaborative meta-analysis of 102 prospective studies. The Lancet, 375(9733), 2215– 2222. https://doi.org/10.1016/s0140-6736(10)60484-9 Schmilovitz-Weiss, H., Gingold-Belfer, R., Boltin, D., Beloosesky, Y., Meyerovitch, J., Tor, R., et al. (2018). Risk of mortality and level of serum alanine aminotransferase among community- dwelling elderly in Israel. European Journal of Gastroenterology and Hepatology, 30(12), 1428–1433. https://doi.org/10.1097/meg.0000000000001225 Schupf, N., Costa, R., Luchsinger, J., Tang, M. X., Lee, J. H., & Mayeux, R. (2005). Relationship between plasma lipids and all-cause mortality in nondemented elderly. Journal of the American Geriatrics Society, 53(2), 219–226. https://doi.org/10.1111/J.1532-5415.2005.53106.x SCORE2 working group and ESC Cardiovascular risk collaboration (2021). SCORE2 risk prediction algorithms: new models to estimate 10-year risk of cardiFovascular disease in Europe. European Heart Journal, 42(25), 2439-2454. https://doi.org/10.1093/eurheartj/ehab309 SCORE2-OP working group and ESC Cardiovascular risk collaboration (2021). SCORE2-OP risk prediction algorithms: estimating incident cardiovascular event risk in older persons in four geographical risk regions. European Heart Journal, 42(25), 2455-2467. https://doi.org/10.1093/eurheartj/ehab312 Searle, S. D., Mitnitski, A., Gahbauer, E. A., Gill, T. M., & Rockwood, K. (2008). A standard procedure for creating a frailty index. BMC Geriatrics, 8, 24. https://doi.org/10.1186/1471-2318-8-24 Shavelle, R. M., Paculdo, D. R., & MacKenzie, R. (2012). Anemia and mortality in older persons: Does the type of anemia affect survival? International Journal of Hematology, 95(3), 248–256. https://doi.org/10.1007/s12185-012-1007-z Sheikh Rezaei, S., Weisshaar, S., Litschauer, B., Gouya, G., Ohrenberger, G., & Wolzt, M. (2019). ADMA and NT pro-BNP are associated with overall mortality in elderly. European Journal of Clinical Investigation, 49(1). https://doi.org/10.1111/eci.13041 Song, X., Mitnitski, A., & Rockwood, K. (2010). Prevalence and 10-year outcomes of frailty in older adults in relation to deficit accumulation. Journal of the American Geriatrics Society, 58(4), 681–687. https://doi.org/10.1111/j.1532-5415.2010.02764.x Sosiaali- ja terveysministeriö. (2024). Laatusuositus aktiivisen ja toimintakykyisen ikääntymisen ja kestävien palvelujen turvaamiseksi 2024–2027. Stamler, J., Stamler, R., Neaton, J. D., Wentworth, D., Daviglus, M. L., Garside, D., Dyer, A. R., Liu, K., & Greenland, P. (1999). Low Risk-Factor Profile and Long-term Cardiovascular and Noncardiovascular Mortality and Life Expectancy: Findings for 5 Large Cohorts of Young Adult and Middle-Aged Men and Women. JAMA, 282(21), 2012–2018. https://doi.org/10.1001/jama.282.21.2012 Strandberg, T., Stenholm, S., Strandberg, A., Salomaa, V., Pitkala, K., & Tilvis, R. (2013). The “Obesity Paradox,” Frailty, Disability, and Mortality in Older Men: A Prospective, Longitudinal Cohort Study. American Journal of Epidemiology, 178(9), 1452–1460. https://doi.org/10.1093/aje/kwt157 References 85 Strandberg, T. E., & Nieminen, T. (2020). Future Perspectives on the Role of Frailty in Cardiovascular Diseases. Advances in Experimental Medicine and Biology, 1216, 149–152. https://doi.org/10.1007/978-3-030-33330-0_14 Suominen, M. H., Jyvakorpi, S. K., Pitkala, K. H., Finne-Soveri, H., Hakala, P., Mannisto, S., Soini, H., Sarlio-Lahteenkorva, S. (2014). Nutritional guidelines for older people in Finland. The Journal of Nutrition, Health & Aging, 18(10), 861–867. https://doi.org/10.1007/s12603-014- 0509-1 Taylor, C. J., Rutten, F. H., Brouwer, J. R., & Hobbs, F. R. (2017). Practical guidance on heart failure diagnosis and management in primary care: recent EPCCS recommendations. Br J Gen Pract, 67(660), 326–327. https://doi.org/10.3399/bjgp17x691553 The Emerging Risk Factors Collaboration (2012). C-Reactive Protein, Fibrinogen, and Cardiovascular Disease Prediction. New England Journal of Medicine, 367(14), 1310–1320. https://doi.org/10.1056/nejmoa1107477 The Global Cardiovascular Risk Consortium. (2025). Global Effect of Cardiovascular Risk Factors on Lifetime Estimates. New England Journal of Medicine. https://doi.org/10.1056/nejmoa2415879 Thompson, M. Q., Theou, O., Tucker, G. R., Adams, R. J., & Visvanathan, R. (2020). FRAIL scale: Predictive validity and diagnostic test accuracy. Australasian Journal on Ageing, 39(4), e529– e536. https://doi.org/10.1111/ajag.12829 Thygesen, K., Alpert, J. S., White, H. D., & Infarction, on behalf of the J. E. T. F. for the R. of M. (2007). Universal Definition of Myocardial Infarction. Journal of the American College of Cardiology, 50(22), 2173–2195. https://doi.org/10.1016/j.jacc.2007.09.011 Thygesen, K., Alpert, J. S., Jaffe, A. S., Simoons, M. L., Chaitman, B. R., White, H. D., et al. (2012). Third universal definition of myocardial infarction. Circulation, 126(16), 2020–2035. https://doi.org/10.1161/cir.0b013e31826e1058 Thygesen, K., Alpert J., Jaffe A., Chaitman, B., Bax, J., Morrow, D., et al. (2018). Fourth Universal Definition of Myocardial Infarction (2018). Journal of the American College of Cardiology, 72(18), 2231–2264. https://doi.org/10.1016/j.jacc.2018.08.1038 Tikhonoff, V., Casiglia, E., Mazza, A., Scarpa, R., Thijs, L., Pessina, A. C., & Staessen, J. A. (2005). Low-density lipoprotein cholesterol and mortality in older people. Journal of the American Geriatrics Society, 53(12), 2159–2164. https://doi.org/10.1111/j.1532-5415.2005.00492.x Tseng, W. C., Chen, Y. T., Ou, S. M., Shih, C. J., Tarng, D. C., Yang, C. Y., Lin, et al. (2018). U- shaped association between serum uric acid levels with cardiovascular and all-cause mortality in the elderly: The role of malnourishment. Journal of the American Heart Association, 7(4). https://doi.org/10.1161/jaha.117.007523 Tzikas, S., Keller, T., Wild, P. S., Schulz, A., Zwiener, I., Zeller, T., et al. (2013). Midregional pro- atrial natriuretic peptide in the general population. Clinical Chemistry and Laboratory Medicine, 51(5), 1125–1133. https://doi.org/10.1515/cclm-2012-0541 United Nations, Department of Economic and Social Affairs, Population Division. World Population Prospects 2024. https://population.un.org/wpp/publications Vargese, S., Jylhä, M., Raitanen, J., Enroth, L., Halonen, P., & Aaltonen, M. (2023). Dementia-related disability in the population aged 90 years and over: differences over time and the role of comorbidity in the vitality 90 + study. BMC Geriatrics, 23(1), 1-11. https://doi.org/10.1186/s12877-023-03980-5 Veronese, N., Cereda, E., Stubbs, B., Solmi, M., Luchini, C., Manzato, et al. (2017). Risk of cardiovascular disease morbidity and mortality in frail and pre-frail older adults: Results from a meta-analysis and exploratory meta-regression analysis. Ageing Research Reviews, 35, 63–73. https://doi.org/10.1016/j.arr.2017.01.003 Vicencio, J. M., Galluzzi, L., Tajeddine, N., Ortiz, C., Criollo, A., Tasdemir, E., Morselli, E., Ben Younes, A., Maiuri, M. C., Lavandero, S., & Kroemer, G. (2008). Senescence, Apoptosis or Autophagy?When a Damaged Cell Must Decide Its Path – A Mini-Review. Gerontology, 54(2), 92–99. https://doi.org/10.1159/000129697 Vilela de Sousa, T., Cavalcante, A. M. R. Z., Lima, N. X., Souza, J. S., Sousa, A. L. L., Brasil, V. V., et al. (2023). Cardiovascular risk factors in the elderly: a 10-year follow-up survival analysis. European Journal of Cardiovascular Nursing, 22(1), 43–52. https://doi.org/10.1093/eurjcn/zvac040 Elisa Heikkilä 86 Viljanen, A., Salminen, M., Irjala, K., Korhonen, P., Wuorela, M., Isoaho, R., Kivelä, S. L., Vahlberg, T., Viitanen, M., Löppönen, M., & Viikari, L. (2021). Frailty, walking ability and self-rated health in predicting institutionalization: an 18-year follow-up study among Finnish community- dwelling older people. Aging Clinical and Experimental Research, 33(3), 547–554. https://doi.org/10.1007/s40520-020-01551-x Visseren, F., Mach, F., Smulders, Y., Carballo, D., Koskinas, K., Bäck, M., et al. (2021). ESC Guidelines on cardiovascular disease prevention in clinical practice: Developed by the Task Force for cardiovascular disease prevention in clinical practice with representatives of the European Society of Cardiology and 12 medical societies with the special contribution of the European Association of Preventive Cardiology (EAPC). European Heart Journal, 42(34), 3227–3337. https://doi.org/10.1093/eurheartj/ehab484 Volpe, M. (2014). Natriuretic peptides and cardio-renal disease. International Journal of Cardiology, 176(3), 630–639. https://doi.org/10.1016/J.ijcard.2014.08.032 von Bonsdorff, M., Rantanen, T., Laukkanen, P., Suutama, T., & Heikkinen, E. (2006). Mobility limitations and cognitive deficits as predictors of institutionalization among community- dwelling older people. Gerontology, 52(6), 359–365. https://doi.org/10.1159/000094985 Waikar, S. S., Mount, D. B., & Curhan, G. C. (2009). Mortality after Hospitalization with Mild, Moderate, and Severe Hyponatremia. American Journal of Medicine, 122(9), 857–865. https://doi.org/10.1016/j.amjmed.2009.01.027 Wang, K., Ni, G., Wu, Q., Zhou, Y., Yao, W., Zhang, H., & Li, X. (2020). Prognostic Value of N-Terminal Pro-B-Type Natriuretic Peptide and Glomerular Filtration Rate in Patients With Acute Heart Failure. Frontiers in Cardiovascular Medicine, 7. https://doi.org/10.3389/fcvm.2020.00123 Wang, L., Lei, J., Wang, R., & Li, K. (2023). Non-Traditional Risk Factors as Contributors to Cardiovascular Disease. Reviews in Cardiovascular Medicine, 24(5), 134. https://doi.org/10.31083/j.rcm2405134 Wang, M., Cao, N., Zhou, L., Su, W., Chen, H., & Li, H. (2023). Association of N-terminal pro-B-type natriuretic peptide levels and mortality risk in acute myocardial infarction across body mass index categories: an observational cohort study. Diabetology and Metabolic Syndrome, 15(1), 1–13. https://doi.org/10.1186/S13098-023-01163-1 Wang, T. J., Larson, M. G. D., Levy, D., Benjamin, E. J., Leip, E. P., Omland, T., Wolf, P. A., Vasan, R. S., & Med, B. (2004). Plasma Natriuretic Peptide Levels and the Risk of Cardiovascular Events and Death From the Framingham Heart Study. N Engl J Med, 7, 655–663. https://doi.org/10.1056/nejmoa031994 Wang, Y., Zhang, R., Shen, Y., Su, L., Dong, B., & Hao, Q. (2019). Prediction of chemotherapy adverse reactions and mortality in older patients with primary lung cancer through frailty index based on routine laboratory data. Clinical Interventions in Aging, 14, 1187–1197. https://doi.org/10.2147/cia.s201873 Wei, Y., Lv, Y., Zhou, J., Gao, X., Duan, J., Zhao, C., Yin, Z., Kang, Q., Wu, B., Chen, C., Mao, C., Li, J., & Shi, X. (2021). Smoking cessation in late life is associated with increased risk of all- cause mortality amongst oldest old people: a community-based prospective cohort study. Age and Ageing, 50(4), 1298–1305. https://doi.org/10.1093/ageing/afaa280 Westermann, D., Neumann, J. T., Sörensen, N. A., & Blankenberg, S. (2017). High-sensitivity assays for troponin in patients with cardiac disease. Nature Reviews Cardiology, 14(8), 472–483. https://doi.org/10.1038/nrcardio.2017.48 Wijkman, M. O., Claggett, B. L., Malachias, M. V. B., Vaduganathan, M., Ballantyne, C. M., Kitzman, D. W., Mosley, T., Matsushita, K., Solomon, S. D., & Pfeffer, M. A. (2022). Importance of NT-proBNP and conventional risk factors for prediction of death in older adults with and without diabetes mellitus– A report from the Atherosclerosis Risk in Communities (ARIC) study. Diabetes Research and Clinical Practice, 194, 110164. https://doi.org/10.1016/j.diabres.2022.110164 Willeit, P., Welsh, P., Evans, J. D. W., Tschiderer, L., Boachie, C., Jukema, J. W., Ford, I., Trompet, S., Stott, D. J., Kearney, P. M., Mooijaart, S. P., Kiechl, S., Di Angelantonio, E., & Sattar, N. (2017). High-Sensitivity Cardiac Troponin Concentration and Risk of First-Ever Cardiovascular Outcomes in 154,052 Participants. Journal of the American College of Cardiology, 70(5), 558– 568. https://doi.org/10.1016/j.jacc.2017.05.062 References 87 Willey, J. Z., Moon, Y. P., Ali Husain, S., Elkind, M. S. V., Sacco, R. L., Wolf, M., Cheung, K., Wright, C. B., & Mohan, S. (2020). Creatinine versus cystatin C for renal function-based mortality prediction in an elderly cohort: The Northern Manhattan Study. PLoS ONE, 15(1), e0226509. https://doi.org/10.1371/journal.pone.0226509 Wilson, P. W. F., D’Agostino, R. B., Levy, D., Belanger, A. M., Silbershatz, H., & Kannel, W. B. (1998). Prediction of coronary heart disease using risk factor categories. Circulation, 97(18), 1837–1847. https://doi.org/10.1161/01.cir.97.18.1837 Wolfgang, C. L. (2013). Role of Fourth-Generation Troponin in Predicting Mortality in Noncardiac Surgery. JAMA Surgery, 148(1), 12–13. https://doi.org/10.1001/jamasurg.2013.607 World Heatlh Organization (2024). Global Health Estimates 2021: Life expectancy and leading causes of death and disability. https://www.who.int/data/gho/data/themes/mortality-and-global-health- estimates Wu, W., Li, D. X., Wang, Q., Xu, Y., & Cui, Y. J. (2018). Relationship between high-sensitivity cardiac troponin T and the prognosis of elderly inpatients with non-acute coronary syndromes. Clinical Interventions in Aging, 13, 1091–1098. https://doi.org/10.2147/cia.s157048 Xu, J., Yu, S., Huang, J., Geng, L., Wei, Y., Li, X., & Gong, X. (2025). Association between admission serum calcium levels and in-hospital mortality in acute myocardial infarction patients with cardiovascular-kidney-metabolic syndrome. Endocrine, 1–8. https://doi.org/10.1007/s12020- 025-04217-8/tables/4 Xu, R. Y., Zhu, X. F., Yang, Y., & Ye, P. (2013). High-sensitive cardiac troponin T. Journal of Geriatric Cardiology, 10(1), 102. https://doi.org/10.3969/j.issn.1671-5411.2013.01.015 Yan, W., Yan, M., Wang, H., & Xu, Z. (2023). Associations of serum alkaline phosphatase level with all-cause and cardiovascular mortality in the general population. Frontiers in Endocrinology, 14. https://doi.org/10.3389/fendo.2023.1217369 Yang, M., Zhuo, Y., Hu, X., & Xie, L. (2018). Predictive validity of two frailty tools for mortality in Chinese nursing home residents: frailty index based on common laboratory tests (FI-Lab) versus FRAIL-NH. Aging Clinical and Experimental Research, 30(12), 1445–1452. https://doi.org/10.1007/s40520-018-1041-7 Yang, P. S., Jang, E., Yu, H. T., Kim, T. H., Pak, H. N., Lee, M. H., & Joung, B. (2021). Changes in cardiovascular risk factors and cardiovascular events in the elderly population. Journal of the American Heart Association, 10(11), 19482. https://doi.org/10.1161/jaha.120.019482 Youden, W. J. (1950). Index for rating diagnostic tests. Cancer. 3(1): 32–35. https://doi.org/10.1002/1097-0142(1950)3:1<32::aid-cncr2820030106>3.0.co;2-3 Zhang, X., Han, X, Zhao, M., Mu, R., Wang, S., Yun, K., & Shang, H. (2020). Determination of high- sensitivity cardiac troponin T upper reference limits under the improved selection criteria in a Chinese population. J Clin Lab Anal, 34. https://doi.org/10.1002/jcla.23007 88 List of Figures and Tables Figures Figure 1. Age structure of population by age group and gender in Finland 31.12.2024 and 2075, projection 2024. .................. 13 Figure 2. Population 1870–2024 and population projection 2025–2075 in Finland by age group, under 15-year- olds indicated by black line, 15 to 64-year-olds by blue line, and 65-year-olds and older by red line. ....................... 13 Figure 3. Over 65-year-old population 2007–2024 and population projection 2025–2075 in Finland by age groups 65–74, 75–84 and 85–. ........................................... 14 Figure 4. Flow chart of the study participants..................................... 39 Figure 5. Survival curves by the laboratory index (LI) comprising 14 commonly used laboratory tests during the 18-year follow-up. ............................................................................ 42 Figure 6. Institutionalization (A) and mortality (B) rates by each index score during the 10-year follow-up. ........................... 46 Figure 7. Rates of institutionalization by the reduced index with the cut-off score ≥4 (A) and mortality by the reduced index with the cut-off score ≥8 (B) during the 10-year follow-up. Age and gender are included in both reduced indexes. ................................................................ 47 Figure 8. The relation of cardiac troponin T (cTnT) in men (A) and women (B), and N-terminal natriuretic b-type propeptide (proBNP) in men (C) and women (D) with age in the reference population. ......................................... 49 Tables Table 1. Reference ranges used for the analytes in the laboratory index for men and women .................................. 34 Table 2. The clinical parameters used in the combined indexes ....... 35 Table 3. Baseline characteristics of study participants for mortality analyses in study I (n=1153)................................. 40 Table 4. Hazard ratios (HR) and their 95% confidence intervals (CI) (in parentheses) of the laboratory index (LI) for mortality during the 10-year follow-up ................................. 41 Table 5. Hazard ratios (HR) and their 95% confidence intervals (CI) (in parentheses) of the laboratory index (LI) for mortality during the 18-year follow-up ................................. 41 List of Figures and Tables 89 Table 6. AUC-values, cut-off limits and their sensitivities, specificities and positive and negative predictive values for the 50-parameter combined index and for the reduced indexes in predicting 10-year institutionalization and mortality ......................................... 44 Table 7. The parameters and their scoring for the reduced indexes .............................................................................. 45 Table 8. Median concentrations and interquartile ranges (IQR) of cardiac troponin T (cTnT) and N-terminal natriuretic b-type propeptide (proBNP) in different age groups among the reference population ........................................ 48 Table 9. Reference limits and their 95% confidence intervals (CI) for cardiac troponin T (cTnT) and N-terminal natriuretic b-type propeptide (proBNP) ............................... 50 Elisa Heikkilä D 1957 AN N ALES UN IVERSITATIS TURKUEN SIS ISBN 978-952-02-0590-4 (PRINT) ISBN 978-952-02-0591-1 (PDF) ISSN 0355-9483 (Print) ISSN 2343-3213 (Online) Pa in os al am a, Tu rk u, F in la nd 2 02 6