| Title: | Age prediction model construction incorporating social determinants of health and its applications to the aging-related disease prediction |
| Journal: | Computational and Structural Biotechnology Reports |
| Published: | 1 Jan 2025 |
| DOI: | https://doi.org/10.1016/j.csbr.2025.100045 |
Publication 15118
WARNING: the interactive features of this website use CSS3, which your browser does not support. To use the full features of this website, please update your browser.
Abstract
Current human age prediction models enhanced aging research by employing diverse biological data, yielding accurate insights into the aging process and associated disease risks. Although it is widely recognized that social determinants of health (SDOH) might affect aging, limited research has incorporated this non-biological variable into age prediction models. We hypothesized that incorporating SDOH into age prediction models could improve the forecasting of aging-related disease risks. This study utilized a comprehensive dataset from the UK Biobank, comprising biological assays, physical measurements, and 10 specific SDOH variables from 343,550 participants of white ancestry, to create a deep neural network-based age prediction model designated SDOHAge, for males and females independently. We computed age acceleration (ΔAge) and employed Cox proportional hazards models to evaluate the association between ΔAge and the incidence of aging-related diseases. The findings demonstrate that the SDOHAge model surpasses models lacking SDOH inputs (C-index values: 0.674 vs. 0.644 for males, and 0.698 vs. 0.654 for females). To explore the biological mechanisms connecting SDOHAge-derived ΔAge with aging, we performed a genome-wide association study and identified multiple gene families, including cystatin (CST), UDP-glucuronosyltransferase 1 A (UGT1A), major histocompatibility complex (MHC)-related genes, and apolipoprotein-related genes. Bidirectional Mendelian randomization further assessed causal associations between ΔAge and aging-related diseases. By delineating the top and bottom 10 % of ΔAge, indicative of the fastest and slowest aging subgroups, we investigated lifestyle factors affecting aging. These findings underscore the influence of SDOH on aging and propose novel avenues for targeted treatments to alleviate health hazards associated with aging.</p>
6 Authors
- Dai-Yuan Liu
- Xiao-Yan Yu
- Rong Huang
- Bei-Hai Tian
- Yuan Quan
- Hong-Yu Zhang
1 Application
| Application ID | Title |
|---|---|
| 97563 | Facilitating combinatorial drug discovery by targeting epistatic disease genes. |
Enabling scientific discoveries that improve human health