Short abstract

Polypharmacy, defined as taking ≥5 different daily medications, is common in older adults and has been linked with neuropsychiatric/neurological and other health conditions. To clarify its potential molecular implications, we tested the hypothesis that polypharmacy may influence DNA methylation (DNAm).

In a longitudinal Italian population cohort - the Moli-sani study (baseline recruitment: 2005-2010, follow-up 2018-2020) - we analyzed DNAm epigenome-wide at each time point for 1,098 participants free from polypharmacy at baseline (mean (SD) age at recruitment: 58.8 (5.6) years, 51.3% women), testing associations of the switch to polypharmacy during follow-up (from 2005-2010 to 2018-2020; median (IQR) 12.6 (1.1) years).with several DNAm aging clocks (Hannum, Horvath, DNAmPhenoAge, DunedinPACE). Then we carried out an epigenome-wide association scan over 668,413 CpGs, testing enrichment of associations for several gene sets and pathways available in public databases.

The analysis of epigenetic clocks revealed a substantial association of DunedinPACE (mean (SD): 1.09 (0.10) years of biological aging per year of chronological aging) with the switch to polypharmacy status (Beta (SE) = 0.0012 (0.0004), p = 0.009). No statistically significant associations were observed at the single CpG level (top hit: cg07675998; chr11q13.1; Beta (SE) = 0.009 (0.002); p = 1.5×10^-6). Epigenome-wide associations showed significant enrichments of several biological functions and pathways related to renal tissue, lipoproteins, inflammatory and immune response.

These findings suggest an influence of polypharmacy on accelerated epigenetic aging and on altered methylation patterns in the genome, suggesting specific pathways as potential targets for mitigating the disruptive effects of polypharmacy on elderly health.

Extended abstract

Background: Polypharmacy, defined as taking five or more different medications, is common in older adults and has been associated with several adverse health outcomes, including mortality, cognitive and functional impairment. Nonetheless, the mechanisms through which this detrimental effect is exerted remain largely unclear. We aimed to test the hypothesis that polypharmacy may influence DNA methylation patterns, in a longitudinal Italian aging cohort.

Methods: We used the Illumina© EPIC array v1 (865,918 CpG sites) to carry out a methylation analysis of a longitudinal subcohort of the Moli-sani study free from polypharmacy at baseline (N=1,098; mean (SD) age at recruitment: 58.8 (5.6) years, 51.3% women). We computed two first generation (Hannum and Horvath), a second (DNAmPhenoAge) and a third generation DNA methylation aging clock (DunedinPACE), and tested them for association with the switch to polypharmacy status during follow-up (from 2005-2010 to 2018-2020; median (IQR) 12.6 (1.1) years). To do so, we used linear mixed effect models adjusted for age, sex, education, prevalent health conditions and lifestyles, in addition to leukocytes counts and residual batch effects. Then we tested associations with 668,413 CpGs passing quality control epigenome-wide. Single CpG associations were used to test enrichment for all the gene sets and pathways available in the Gene Ontology (GO) terms, in the REACTOME and in the Kyoto Encyclopedia of Genes and Genomes (KEGG) databases.

Results: The analysis of DNAm aging clocks revealed a small but significant association of DunedinPACE (mean (SD): 1.09 (0.10) year of biological aging per year of chronological aging) with the switch to polypharmacy status during follow-up (Beta (SE) = 0.0012 (0.0004), p = 0.009). The epigenome-wide association scan with polypharmacy showed no significant associations. The top hit was detected at cg07675998 (chr11q13.1; Beta (SE) = 0.009 (0.002); p = 1.5×10-6), a site previously associated with circulating C-reactive protein levels. A pathway analysis revealed significant enrichments of CpG associations (FDR <0.05) for several Gene Ontology biological functions and REACTOME pathways, mostly related to renal tissue development, lipoproteins and cholesterol homeostasis and biosynthesis, inflammatory and immune response.

Conclusions: This represents the first evidence of an influence of polypharmacy on accelerated epigenetic aging and altered methylation patterns in the genome. These novel findings are consistent with a previous proteomic analysis of liver tissues in aging mouse models under polypharmacy and suggest potential targets for mitigating disruptive effects of polypharmacy on elderly health.

In aging populations, one key factor to prevent adverse health outcomes and allocate healthcare resources is the possibility of identifying precise multimorbidity patterns and seizing their trajectories in time. Using longitudinal data of individuals aged 50 and above residing Emilia-Romagna region in 2011 and followed up to 2019 (N = 1,010,571), we document the multimorbidity patterns transition leading to death at older ages across three time-points (in 2011, 2016, 2019). To achieve this objective, our analysis is structured in three steps, corresponding to three questions: (i) what are the multimorbidity patterns in older-aged population? (ii) how does multimorbidity patterns transition can be quantified in data at individual level? (iii) how can we model this transition in patterns across time to extract information most valuable for policy planning? First, we combine graphical model and network analysis to identify multimorbidity patterns at each time-point. Second, we introduce a scoring method using network metrics to assign unique multimorbidity pattern to each individual at each time-point. Third, we apply hidden Markov model to capture the transition process between multimorbidity patterns at older ages, accounting for latent factor that could make the observed transition differs from the true underlying process. We stratify our population by sex and age groups (50-59, 60-69, 70-79, 80+) and conduct the three-steps analyses systematically on 8 subgroups. Thereby, we demonstrate how multimorbidity at older ages can be studied under network perspective, and offer in-depth understanding on the evolution of multimorbidity at older ages for males and females at different ages.

Providing sustainable and quality care to ageing population requires improving and maintaining care pathways performance, in terms of increased efficiency, timeliness, effectiveness, patient safety, accessibility and care integration. To this end, Lean and Safety Management constitutes a managerial approach to maximize patient value, by reducing process wastes and risks. The current research broadens the scope of this approach to the design of new sustainable care pathways, with the goal of optimizing their performance from the outset. A real-world case study was developed, focusing on the design of a new self-guided home-based upper limb rehabilitation pathway for post-stroke patients. The new pathway was developed starting from the design of the clinical activities and then optimizing the performance by analyzing the organizational and managerial aspects. The results demonstrate enhanced effectiveness and timeliness of care, service accessibility, productivity, and efficiency for the patient, the caregiver and care provider. The research offers a practical contribution by supporting healthcare management and care providers in enhancing healthcare sustainability, a social contribution by improving the quality of care and life for patients and caregivers, and an academic contribution by extending the typical scope of application of Lean and Safety Management approach.

Background: Aging represents a complex and heterogeneous process characterized by an irreversible and progressive decline in physiological integrity, compromising organ function, and increasing vulnerability to chronic diseases. While environmental factors contribute to disease development, aging remains the most important risk factor for morbidity and mortality. Chronological age (ChronoAge) poorly reflects individual health variability, bypassing environmental and biological influences that could shape aging trajectories. In contrast, phenotypic age (PhenoAge), a biomarker-based estimate of biological age, provides a more accurate reflection of individual aging and demonstrates superior predictive power for disease onset and mortality. A central hallmark of aging is inflammaging, a chronic, low-grade inflammatory state sustained by processes like immune cell activation and cellular senescence. Elevated levels of inflammatory markers have been linked to several age-related diseases and adverse outcomes, including frailty, hospitalization, and death. Therefore, profiling inflammation across the aging trajectory may reveal key biological mechanisms and risk signatures linked to accelerated aging and age-associated diseases.

Material and methods: Participants in this study were enrolled in the Novara Cohort Study (NCS), a population-based longitudinal and cross-sectional study designed to investigate the biological, behavioral, and psychosocial determinants of aging in adults residing in the Novara province (Italy). The NCS integrates biological sample collection with comprehensive assessments of lifestyle, clinical status, and physical and cognitive function to support biomarker discovery and aging trajectory modeling. Proteomic profiling with the Olink Target 96 Inflammation panel is currently being analyzed on plasma EDTA samples from 500 participants. PhenoAge was calculated for all subjects based on the estimation proposed by Levine. Inflammatory proteins correlated with age and health-related parameters through Spearman’s rank correlation coefficient. Variable Importance in Projection (VIP) analysis from orthogonal PLS-DA models was used to identify proteins discriminating between age classes and aging phenotypes. Functional enrichment analysis was performed to determine underlying biological processes.

Results: Participants included in this study showed a significantly lower phenotypic age compared to chronological age (mean PhenoAccell=-7.6 years). Both chronological age and phenotypic age showed a significant positive correlation with BMI and clinical frailty scale (CFS), while they were negatively correlated with cognitive performance (assessed by the Montreal Cognitive Assessment, MoCA) and physical function (SPPB). Targeted proteomic profiling of inflammatory proteins revealed 24 markers that were significantly and positively correlated with chronological and phenotypic age. Among the strongest chronological age-associated inflammatory proteins were TNFRSF9 (ρ = 0.664), CXCL9 (ρ = 0.654), and CDCP1 (ρ = 0.643). PhenoAge showed slightly weaker, yet still significant, correlation with CDCP1 (ρ = 0.519), OPG (ρ = 0.495), and CXCL9 (ρ = 0.494). Additionally, 34 proteins showed a significant positive correlation with PhenoAccel, whereas two proteins, uPA and TWEAK, were negatively correlated with PhenoAccel.

Conclusion: Although inflammation-related proteins tend to show stronger correlations with chronological age, their expression levels increase more markedly in biologically older individuals, reinforcing the value of PhenoAge as a marker of inflammaging. These findings highlight the importance of integrating biological age metrics into aging research and support using proteomic signatures to monitor aging trajectories. Future studies with longitudinal follow-up will help determine whether these inflammatory profiles predict age-related decline and serve as potential targets for preventive interventions.

SHORT ABSTRACT

Aging is a multifaceted process marked by a gradual decline in physiological integrity, increasing susceptibility to chronic diseases. While chronological age (ChronoAge) has traditionally defined aging, it often fails to reflect interindividual variability in health. In contrast, phenotypic age (PhenoAge), a biomarker-derived estimate of biological age, offers a more precise prediction of morbidity and mortality. A key contributor to aging is inflammaging, a chronic low-grade inflammatory state associated with cellular senescence and immune activation. In this study, we investigated inflammatory protein signatures associated with aging and age-related outcomes in the Novara Cohort Study, a longitudinal population study coordinated by UPO. Participants underwent comprehensive clinical and functional assessments, and plasma samples from 500 participants are being analyzed using the Olink Target 96 Inflammation panel. PhenoAge and PhenoAccel were computed based on established algorithms. NCS participants exhibited a significantly lower PhenoAge than ChronoAge (mean PhenoAccel = –7.6 years). Both age metrics correlated positively with BMI and frailty, and negatively with cognitive and physical performance. Proteomic analysis identified 24 inflammatory proteins positively correlated with both ChronoAge and PhenoAge, with stronger correlations generally observed with ChronoAge. Notable proteins included TNFRSF9, CXCL9, and CDCP1. Additionally, 34 proteins positively correlated with PhenoAccel, while two (uPA and TWEAK) were negatively associated. These findings underscore the role of inflammation in biological aging and support the utility of PhenoAge as a marker of inflammaging. Integrating proteomic data with biological age estimation may improve understanding of aging mechanisms and inform preventive strategies targeting age-related decline.