Genetics and biomarkers of frailty : towards individualized management of the frailty syndrome

Abstract: Frailty is an age-related, dynamic state of multisystem physiological decline and is a strong predictor of disability and mortality. To move towards an individualized management of frailty, a better understanding of its biological underpinnings and an early identification of frail older adults are necessary. The overarching aim of this thesis was to unravel the genetic, epigenetic, and metabolic determinants of frailty and to develop an automated frailty assessment tool for the Swedish health system. In Study I, we calculated a frailty index (FI) for 42,994 Swedish twins and assessed sex differences in the genetic and environmental contributions to the FI. Overall, we observed a higher heritability of the FI in women (52%) than in men (45%). Moreover, the correlations between FI and its two main risk factors, body mass index and education, were mainly attributable to genetic factors and environmental factors shared within twin pairs, respectively, suggesting that different mechanisms may underlie these associations. In Study II, we examined genetic and environmental influences on the FI trajectories in 2,496 twins followed up to 27 years. A bilinear latent growth curve model best fit the data, indicating a four-to-five times faster FI increase after age 75. While genetic influences were relatively stable across age, individual-specific environmental influences increased substantially after age 75 especially in men, amplifying the overall FI variance in late life. In Study III, we performed an epigenome-wide analysis in 526 Swedish twins and identified 171 CpG sites associated with the FI at a false discovery rate of <0.05. Many of the identified sites have previously been associated with chronological age and age-related diseases. We further validated five of these sites in an independent sample of 304 Danish twins, which are mapped to genes that may involve in cancer and neurological pathways. In Study IV, we explored the associations of 168 metabolomic and 32 clinical biomarkers with two measures of frailty using observational and Mendelian randomization analyses. In three population-based studies comprising >100,000 individuals, we identified 34 biomarkers independently and robustly associated with the FI. Specifically, we highlighted a putative causal effect of glycoprotein acetyls, an inflammatory biomarker, on frailty. In Study V, we developed an electronic frailty index (eFI) using electronic health records from 18,225 patients admitted to nine geriatric clinics in Stockholm. Among the assessed frailty and comorbidity measures, the eFI had the best discriminative ability for mortality. In summary, this thesis provides novel insights into the biological mechanisms of frailty, suggesting that both genetic and environmental factors play important roles in frailty development, with chronic inflammation as the key underlying mechanism. Moreover, our developed Swedish eFI is a promising tool that can potentially be incorporated in the Swedish health system to guide clinical decisions.