Thèse Biologie Intégrative des Systèmes du Vieillissement en Bonne Santé une Signature Sanguine pour Prédire l'Évolution Longitudinale de la Capacité Intrinsèque H/F Doctorat.Gouv.Fr

Établissement : Université de Toulouse École doctorale : BSB - Biologie, Santé, Biotechnologies Laboratoire de recherche : RESTORE Direction de la thèse : Isabelle ADER-PERARNAU ORCID 0000000151038509 Début de la thèse : 2027-09-01 Date limite de candidature : 2026-11-23T23:59:59 Le vieillissement s'accompagne d'une diminution progressive des réserves physiologiques et fonctionnelles, souvent plusieurs années avant l'apparition d'une incapacité ou d'une dépendance. L'Organisation mondiale de la Santé a proposé le concept de capacité intrinsèque, qui regroupe cinq dimensions majeures de la santé fonctionnelle : locomotion, cognition, fonctions sensorielles, vitalité et santé psychologique. Malgré sa pertinence clinique, aucun biomarqueur sanguin n'est aujourd'hui validé pour identifier précocement les personnes à risque de déclin de cette capacité.
Ce projet vise à déterminer si une signature biologique sanguine mesurée à l'inclusion peut prédire les trajectoires ultérieures de capacité intrinsèque au cours du vieillissement. Il s'appuie sur la cohorte INSPIRE-T, qui comprend environ 1 000 volontaires âgés de 20 à plus de 100 ans. Évalués cliniquement chaque année depuis 2019, ils seront suivis pendant douze ans. Les participants inclus dès le début de la cohorte disposent aujourd'hui de six années de suivi, avec une évaluation répétée de leur capacité intrinsèque.
Les données biologiques étudiées correspondent exclusivement à l'état initial des participants. Elles sont déjà disponibles ou seront mesurées au cours du projet à partir des échantillons plasmatiques collectés à l'inclusion. Le projet mettra ainsi en relation une photographie biologique initiale avec l'évolution clinique et fonctionnelle observée annuellement. Cette conception prospective et longitudinale constitue sa principale force.
Les biomarqueurs seront sélectionnés dans un cadre intégratif associant métabolisme énergétique, homéostasie redox, intégrité et remodelage des structures tissulaires et inflammation systémique, sans privilégier une cible moléculaire unique. L'objectif n'est pas de développer un marqueur de l'âge chronologique, mais une signature de l'état physiologique initial capable d'anticiper le maintien ou le déclin des capacités fonctionnelles.
La complexité du vieillissement fonctionnel impose une approche multidisciplinaire. Aucun champ ne peut, à lui seul, relier les mécanismes biologiques systémiques, les données cliniques longitudinales et les trajectoires individuelles de santé. Le projet articule donc physiologie, biologie du vieillissement, géroscience, épidémiologie, biologie des systèmes, biostatistique, modélisation mathématique et science des données. Cette complémentarité est indispensable pour transformer des données biologiques et cliniques hétérogènes en signatures prédictives robustes, interprétables et cliniquement pertinentes.
Les trajectoires globales et spécifiques de capacité intrinsèque seront reconstruites à partir des visites annuelles. Les analyses combineront modèles mixtes, modèles de trajectoires, indices multivariés de dérégulation physiologique, réduction de dimension et apprentissage statistique. Elles évalueront la capacité des données biologiques initiales à prédire le niveau futur de capacité intrinsèque, sa vitesse d'évolution, d'éventuelles ruptures et l'appartenance à des profils de vieillissement fonctionnel défavorables. Les modèles tiendront compte du niveau fonctionnel initial, de l'âge, du sexe et des principaux facteurs de confusion.
Les résultats attendus sont l'identification d'une signature plasmatique pronostique du vieillissement fonctionnel, la détection précoce des individus à risque et une meilleure compréhension des dérégulations systémiques précédant le déclin clinique. Ce travail pourrait poser les bases d'un outil sanguin peu invasif, complémentaire de l'évaluation clinique, et soutenir des stratégies personnalisées de prévention visant à préserver la capacité intrinsèque, l'autonomie et le bien-être. Il illustrera aussi la nécessité d'intégrer les expertises biologiques, cliniques, épidémiologiques et quantitatives pour promouvoir le vieillissement en bonne santé. The increase in life expectancy has not been accompanied by an equivalent increase in the number of years lived in good health. The decline in physiological reserves often begins silently, several years before the onset of frailty, disability, or dependency. Identifying the biological alterations that precede this decline is therefore a major challenge for developing earlier and more personalised prevention strategies.
Within its healthy ageing framework, the World Health Organization introduced the concept of intrinsic capacity, defined as the combination of all the physical and mental capacities available to an individual. Intrinsic capacity encompasses five domains: locomotion, cognition, sensory functions, vitality, and psychological health. The intrinsic capacity score therefore provides an integrated measure of an individual's functional health status and offers a person- and function-centred approach that complements disease-based descriptions. Importantly, a low intrinsic capacity score is associated not only with poorer functional health, but also with a greater burden of comorbidities, frailty, and adverse health outcomes.
The INSPIRE-T cohort provides a particularly suitable framework for investigating these mechanisms. It includes approximately 1,000 adults aged from 20 to over 100 years. Recruitment began in 2019. At the baseline visit, participants underwent clinical phenotyping and plasma sampling. Since enrolment, their intrinsic capacity has been assessed annually. The earliest participants therefore currently have up to six years of clinical follow-up after the initial biological assessment. The duration of available follow-up naturally varies according to each participant's date of inclusion.
This design makes it possible to clearly separate the timing of the predictive measurement from that of the clinical outcome. All biological data analysed in the project correspond to baseline and provide a snapshot of the participants' initial systemic state. In contrast, the repeated clinical data describe the subsequent evolution of functional health. The project will therefore seek prognostic markers that precede decline, rather than merely identifying markers associated with an already impaired functional state.
Our gerophysiological framework is based on three fundamental and interdependent dimensions: tissue structure, metabolism, and inflammation-immunity. Structure refers to tissue integrity, mechanical resilience, and repair capacity. Metabolism reflects the efficiency and adaptability of energy pathways and redox homeostasis. Inflammation-immunity refers to the balance between responses required for repair and chronic low-grade activation that may progressively impair tissues and their function.
Our recent work on primary fibroblasts derived from participants in the INSPIRE-T cohort showed that integrating several cellular functions enables the identification of profiles associated with age, frailty, and intrinsic capacity. At the systemic level, preliminary data also show an association between a global circulating indicator of redox balance and intrinsic capacity. These findings support the hypothesis that a multifactorial plasma signature measured at baseline could reveal latent biological vulnerability and predict subsequent functional outcomes.

The project therefore aims to move beyond a molecule-by-molecule approach. It will assess whether combining functional families of biomarkers with an overall quantification of physiological dysregulation provides more robust, interpretable, and predictive information on intrinsic capacity trajectories. Overall Objective: To determine whether a plasma biological signature measured at baseline can predict subsequent annual intrinsic capacity trajectories, with follow-up currently extending up to six years for the earliest participants in the INSPIRE-T cohort (1,000 men and women aged 20 to 100 years).

Specific Objectives: To build a harmonised biological database exclusively reflecting the participants' baseline status, using both measurements already available and complementary assays that will be performed on plasma samples collected at inclusion.
To harmonise annual clinical assessments and model both global and domain-specific intrinsic capacity trajectories from baseline to the most recent available follow-up visit.
To develop composite biological signatures reflecting the metabolic and redox, structural, and inflammatory dimensions of physiological dysregulation.
To determine whether these baseline signatures predict future intrinsic capacity levels, rates of change, changes in slope, and adverse functional ageing profiles.
To compare the prognostic value of integrative signatures with that of individual biomarkers and with that of clinical variables available at baseline.
To assess the added value of the biological models after adjustment for baseline intrinsic capacity, age, sex, comorbidities, and the main potential confounding factors.
To identify biological profiles that may improve the early detection of individuals at risk, before the onset of clinically apparent functional decline. The project will be based on the prospective INSPIRE-T cohort. The analysis will strictly distinguish between two temporal levels:
- Predictive biological time point: plasma samples and biological data measured exclusively at inclusion (T0).
- Longitudinal clinical time frame: annual assessments of intrinsic capacity from inclusion onwards, with up to six years of follow-up currently available for the earliest participants enrolled in 2019.
The main contribution of the project will be to use the initial biological state to predict both the level and the dynamics of functional health observed over the subsequent years.

WP1: Establishment of the Baseline Biological Database
An inventory of baseline plasma samples and biological data already available will be conducted. Both pre-existing measurements and complementary assays will be generated or analysed exclusively from samples collected during the baseline visit. No biomarker measured during follow-up visits will be included in the main predictive signature.
The biomarker panel will be organised according to functional biological families, without prioritising any single molecular target a priori:
- energy metabolism, metabolic stress responses, and redox homeostasis;
- integrity, repair, and remodelling of tissue structures and the extracellular matrix; endothelial function, cellular interactions, and vascular remodelling;
- systemic inflammation and immune regulation.
Complementary assays will be performed on stored plasma samples using automated and standardised technologies, including multiplex immunoassays, in collaboration with partner platforms. Procedures will include quality-control measures, assessment of batch effects, normalisation, analysis of outliers, and appropriate management of missing data.
Deliverables: inventory of baseline resources; harmonised biological database; data dictionary; quality-control report; and classification of biomarkers according to the functional dimensions of the project.

WP2: Reconstruction of Longitudinal Intrinsic Capacity Trajectories
Clinical data collected annually will be harmonised from inclusion to the most recent available follow-up visit. Global intrinsic capacity and its five domains will be calculated using consistent procedures across all assessment time points. The data structure will account for the staggered enrolment of participants and, consequently, for differences in individual follow-up duration.
Trajectories will be investigated using linear or non-linear mixed-effects models, supplemented where necessary by latent class models or functional data analysis approaches. Derived indicators will include:
- intrinsic capacity level at baseline;
- intrinsic capacity level observed at each annual follow-up visit;
- mean rate of change and individual deviations from this average slope;
- changes in slope, accelerations, or critical transitions;
- stable, intermediate, or adverse trajectory profiles;
- domain-specific trajectories for each of the five intrinsic capacity domains.
Deliverables: harmonised longitudinal clinical database; trajectory indicators; potential classification of functional ageing profiles; and documentation of calculation rules and missing-data management procedures.

WP3: Integrative Modelling and Prediction
Baseline biomarkers will first be analysed individually and then integrated into composite signatures. Several complementary strategies will be compared, including multivariate measures of physiological dysregulation, principal component analysis, factor models, regularisation and variable-selection methods, supervised learning, and, if data size and quality allow, non-linear representation-learning models.
Three distinct questions will be addressed:
- Cross-sectional association: which baseline biomarkers are associated with intrinsic capacity measured at the same time point?
- Prospective prediction: which baseline biomarkers predict intrinsic capacity levels at subsequent follow-up visits?
- Trajectory prediction: which baseline biological profiles predict the slope of change, trajectory shifts, or membership in a declining trajectory profile ?
Prospective models will be systematically adjusted for baseline intrinsic capacity. This step will make it possible to assess genuine prognostic value, independently of functional differences already present at the time of sample collection. Analyses will also account for age, sex, comorbidities, treatments, and other relevant confounding factors available within the cohort.